k

.github/workflows/pr-chromatic-tests.yml

@@ -8,8 +8,6 @@ on: push
 env:
   OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
   SLACK_BOT_TOKEN: ${{ secrets.SLACK_BOT_TOKEN }}
-  GEN_AI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
-  MOCK_LLM_RESPONSE: true
 
 jobs:
   playwright-tests:

.github/workflows/pr-helm-chart-testing.yml

@@ -21,10 +21,10 @@ jobs:
     - name: Set up Helm
       uses: azure/setup-helm@v4.2.0
       with:
-        version: v3.17.0
+        version: v3.14.4
       
     - name: Set up chart-testing
-      uses: helm/chart-testing-action@v2.7.0
+      uses: helm/chart-testing-action@v2.6.1
 
     # even though we specify chart-dirs in ct.yaml, it isn't used by ct for the list-changed command...
     - name: Run chart-testing (list-changed)
@@ -37,6 +37,22 @@ jobs:
           echo "changed=true" >> "$GITHUB_OUTPUT"
         fi
 
+#     rkuo: I don't think we need python?
+#     - name: Set up Python
+#       uses: actions/setup-python@v5
+#       with:
+#         python-version: '3.11'
+#         cache: 'pip'
+#         cache-dependency-path: |
+#           backend/requirements/default.txt
+#           backend/requirements/dev.txt
+#           backend/requirements/model_server.txt
+#     - run: |
+#         python -m pip install --upgrade pip
+#         pip install --retries 5 --timeout 30 -r backend/requirements/default.txt
+#         pip install --retries 5 --timeout 30 -r backend/requirements/dev.txt
+#         pip install --retries 5 --timeout 30 -r backend/requirements/model_server.txt
+
     # lint all charts if any changes were detected
     - name: Run chart-testing (lint)
       if: steps.list-changed.outputs.changed == 'true'
@@ -46,7 +62,7 @@ jobs:
 
     - name: Create kind cluster
       if: steps.list-changed.outputs.changed == 'true'
-      uses: helm/kind-action@v1.12.0
+      uses: helm/kind-action@v1.10.0
 
     - name: Run chart-testing (install)
       if: steps.list-changed.outputs.changed == 'true'

backend/alembic/versions/2f80c6a2550f_add_chat_session_specific_temperature_.py

@@ -1,36 +0,0 @@
-"""add chat session specific temperature override
-
-Revision ID: 2f80c6a2550f
-Revises: 33ea50e88f24
-Create Date: 2025-01-31 10:30:27.289646
-
-"""
-from alembic import op
-import sqlalchemy as sa
-
-
-# revision identifiers, used by Alembic.
-revision = "2f80c6a2550f"
-down_revision = "33ea50e88f24"
-branch_labels = None
-depends_on = None
-
-
-def upgrade() -> None:
-    op.add_column(
-        "chat_session", sa.Column("temperature_override", sa.Float(), nullable=True)
-    )
-    op.add_column(
-        "user",
-        sa.Column(
-            "temperature_override_enabled",
-            sa.Boolean(),
-            nullable=False,
-            server_default=sa.false(),
-        ),
-    )
-
-
-def downgrade() -> None:
-    op.drop_column("chat_session", "temperature_override")
-    op.drop_column("user", "temperature_override_enabled")

backend/alembic/versions/33ea50e88f24_foreign_key_input_prompts.py

@@ -1,80 +0,0 @@
-"""foreign key input prompts
-
-Revision ID: 33ea50e88f24
-Revises: a6df6b88ef81
-Create Date: 2025-01-29 10:54:22.141765
-
-"""
-from alembic import op
-
-
-# revision identifiers, used by Alembic.
-revision = "33ea50e88f24"
-down_revision = "a6df6b88ef81"
-branch_labels = None
-depends_on = None
-
-
-def upgrade() -> None:
-    # Safely drop constraints if exists
-    op.execute(
-        """
-        ALTER TABLE inputprompt__user
-        DROP CONSTRAINT IF EXISTS inputprompt__user_input_prompt_id_fkey
-        """
-    )
-    op.execute(
-        """
-        ALTER TABLE inputprompt__user
-        DROP CONSTRAINT IF EXISTS inputprompt__user_user_id_fkey
-        """
-    )
-
-    # Recreate with ON DELETE CASCADE
-    op.create_foreign_key(
-        "inputprompt__user_input_prompt_id_fkey",
-        "inputprompt__user",
-        "inputprompt",
-        ["input_prompt_id"],
-        ["id"],
-        ondelete="CASCADE",
-    )
-
-    op.create_foreign_key(
-        "inputprompt__user_user_id_fkey",
-        "inputprompt__user",
-        "user",
-        ["user_id"],
-        ["id"],
-        ondelete="CASCADE",
-    )
-
-
-def downgrade() -> None:
-    # Drop the new FKs with ondelete
-    op.drop_constraint(
-        "inputprompt__user_input_prompt_id_fkey",
-        "inputprompt__user",
-        type_="foreignkey",
-    )
-    op.drop_constraint(
-        "inputprompt__user_user_id_fkey",
-        "inputprompt__user",
-        type_="foreignkey",
-    )
-
-    # Recreate them without cascading
-    op.create_foreign_key(
-        "inputprompt__user_input_prompt_id_fkey",
-        "inputprompt__user",
-        "inputprompt",
-        ["input_prompt_id"],
-        ["id"],
-    )
-    op.create_foreign_key(
-        "inputprompt__user_user_id_fkey",
-        "inputprompt__user",
-        "user",
-        ["user_id"],
-        ["id"],
-    )

backend/alembic/versions/a6df6b88ef81_remove_recent_assistants.py

@@ -1,29 +0,0 @@
-"""remove recent assistants
-
-Revision ID: a6df6b88ef81
-Revises: 4d58345da04a
-Create Date: 2025-01-29 10:25:52.790407
-
-"""
-from alembic import op
-import sqlalchemy as sa
-from sqlalchemy.dialects import postgresql
-
-# revision identifiers, used by Alembic.
-revision = "a6df6b88ef81"
-down_revision = "4d58345da04a"
-branch_labels = None
-depends_on = None
-
-
-def upgrade() -> None:
-    op.drop_column("user", "recent_assistants")
-
-
-def downgrade() -> None:
-    op.add_column(
-        "user",
-        sa.Column(
-            "recent_assistants", postgresql.JSONB(), server_default="[]", nullable=False
-        ),
-    )

backend/ee/onyx/external_permissions/confluence/doc_sync.py

@@ -13,7 +13,6 @@ from onyx.connectors.confluence.onyx_confluence import OnyxConfluence
 from onyx.connectors.confluence.utils import get_user_email_from_username__server
 from onyx.connectors.models import SlimDocument
 from onyx.db.models import ConnectorCredentialPair
-from onyx.indexing.indexing_heartbeat import IndexingHeartbeatInterface
 from onyx.utils.logger import setup_logger
 
 logger = setup_logger()
@@ -258,7 +257,6 @@ def _fetch_all_page_restrictions(
     slim_docs: list[SlimDocument],
     space_permissions_by_space_key: dict[str, ExternalAccess],
     is_cloud: bool,
-    callback: IndexingHeartbeatInterface | None,
 ) -> list[DocExternalAccess]:
     """
     For all pages, if a page has restrictions, then use those restrictions.
@@ -267,12 +265,6 @@ def _fetch_all_page_restrictions(
     document_restrictions: list[DocExternalAccess] = []
 
     for slim_doc in slim_docs:
-        if callback:
-            if callback.should_stop():
-                raise RuntimeError("confluence_doc_sync: Stop signal detected")
-
-            callback.progress("confluence_doc_sync:fetch_all_page_restrictions", 1)
-
         if slim_doc.perm_sync_data is None:
             raise ValueError(
                 f"No permission sync data found for document {slim_doc.id}"
@@ -342,7 +334,7 @@ def _fetch_all_page_restrictions(
 
 
 def confluence_doc_sync(
-    cc_pair: ConnectorCredentialPair, callback: IndexingHeartbeatInterface | None
+    cc_pair: ConnectorCredentialPair,
 ) -> list[DocExternalAccess]:
     """
     Adds the external permissions to the documents in postgres
@@ -367,12 +359,6 @@ def confluence_doc_sync(
     logger.debug("Fetching all slim documents from confluence")
     for doc_batch in confluence_connector.retrieve_all_slim_documents():
         logger.debug(f"Got {len(doc_batch)} slim documents from confluence")
-        if callback:
-            if callback.should_stop():
-                raise RuntimeError("confluence_doc_sync: Stop signal detected")
-
-            callback.progress("confluence_doc_sync", 1)
-
         slim_docs.extend(doc_batch)
 
     logger.debug("Fetching all page restrictions for space")
@@ -381,5 +367,4 @@ def confluence_doc_sync(
         slim_docs=slim_docs,
         space_permissions_by_space_key=space_permissions_by_space_key,
         is_cloud=is_cloud,
-        callback=callback,
     )

backend/ee/onyx/external_permissions/confluence/group_sync.py

@@ -14,8 +14,6 @@ def _build_group_member_email_map(
 ) -> dict[str, set[str]]:
     group_member_emails: dict[str, set[str]] = {}
     for user_result in confluence_client.paginated_cql_user_retrieval():
-        logger.debug(f"Processing groups for user: {user_result}")
-
         user = user_result.get("user", {})
         if not user:
             logger.warning(f"user result missing user field: {user_result}")
@@ -35,17 +33,10 @@ def _build_group_member_email_map(
             logger.warning(f"user result missing email field: {user_result}")
             continue
 
-        all_users_groups: set[str] = set()
         for group in confluence_client.paginated_groups_by_user_retrieval(user):
             # group name uniqueness is enforced by Confluence, so we can use it as a group ID
             group_id = group["name"]
             group_member_emails.setdefault(group_id, set()).add(email)
-            all_users_groups.add(group_id)
-
-        if not group_member_emails:
-            logger.warning(f"No groups found for user with email: {email}")
-        else:
-            logger.debug(f"Found groups {all_users_groups} for user with email {email}")
 
     return group_member_emails
 

backend/ee/onyx/external_permissions/gmail/doc_sync.py

@@ -6,7 +6,6 @@ from onyx.access.models import ExternalAccess
 from onyx.connectors.gmail.connector import GmailConnector
 from onyx.connectors.interfaces import GenerateSlimDocumentOutput
 from onyx.db.models import ConnectorCredentialPair
-from onyx.indexing.indexing_heartbeat import IndexingHeartbeatInterface
 from onyx.utils.logger import setup_logger
 
 logger = setup_logger()
@@ -29,7 +28,7 @@ def _get_slim_doc_generator(
 
 
 def gmail_doc_sync(
-    cc_pair: ConnectorCredentialPair, callback: IndexingHeartbeatInterface | None
+    cc_pair: ConnectorCredentialPair,
 ) -> list[DocExternalAccess]:
     """
     Adds the external permissions to the documents in postgres
@@ -45,12 +44,6 @@ def gmail_doc_sync(
     document_external_access: list[DocExternalAccess] = []
     for slim_doc_batch in slim_doc_generator:
         for slim_doc in slim_doc_batch:
-            if callback:
-                if callback.should_stop():
-                    raise RuntimeError("gmail_doc_sync: Stop signal detected")
-
-                callback.progress("gmail_doc_sync", 1)
-
             if slim_doc.perm_sync_data is None:
                 logger.warning(f"No permissions found for document {slim_doc.id}")
                 continue

backend/ee/onyx/external_permissions/google_drive/doc_sync.py

@@ -10,7 +10,6 @@ from onyx.connectors.google_utils.resources import get_drive_service
 from onyx.connectors.interfaces import GenerateSlimDocumentOutput
 from onyx.connectors.models import SlimDocument
 from onyx.db.models import ConnectorCredentialPair
-from onyx.indexing.indexing_heartbeat import IndexingHeartbeatInterface
 from onyx.utils.logger import setup_logger
 
 logger = setup_logger()
@@ -129,7 +128,7 @@ def _get_permissions_from_slim_doc(
 
 
 def gdrive_doc_sync(
-    cc_pair: ConnectorCredentialPair, callback: IndexingHeartbeatInterface | None
+    cc_pair: ConnectorCredentialPair,
 ) -> list[DocExternalAccess]:
     """
     Adds the external permissions to the documents in postgres
@@ -147,12 +146,6 @@ def gdrive_doc_sync(
     document_external_accesses = []
     for slim_doc_batch in slim_doc_generator:
         for slim_doc in slim_doc_batch:
-            if callback:
-                if callback.should_stop():
-                    raise RuntimeError("gdrive_doc_sync: Stop signal detected")
-
-                callback.progress("gdrive_doc_sync", 1)
-
             ext_access = _get_permissions_from_slim_doc(
                 google_drive_connector=google_drive_connector,
                 slim_doc=slim_doc,

backend/ee/onyx/external_permissions/slack/doc_sync.py

@@ -7,7 +7,6 @@ from onyx.connectors.slack.connector import get_channels
 from onyx.connectors.slack.connector import make_paginated_slack_api_call_w_retries
 from onyx.connectors.slack.connector import SlackPollConnector
 from onyx.db.models import ConnectorCredentialPair
-from onyx.indexing.indexing_heartbeat import IndexingHeartbeatInterface
 from onyx.utils.logger import setup_logger
 
 
@@ -15,7 +14,7 @@ logger = setup_logger()
 
 
 def _get_slack_document_ids_and_channels(
-    cc_pair: ConnectorCredentialPair, callback: IndexingHeartbeatInterface | None
+    cc_pair: ConnectorCredentialPair,
 ) -> dict[str, list[str]]:
     slack_connector = SlackPollConnector(**cc_pair.connector.connector_specific_config)
     slack_connector.load_credentials(cc_pair.credential.credential_json)
@@ -25,14 +24,6 @@ def _get_slack_document_ids_and_channels(
     channel_doc_map: dict[str, list[str]] = {}
     for doc_metadata_batch in slim_doc_generator:
         for doc_metadata in doc_metadata_batch:
-            if callback:
-                if callback.should_stop():
-                    raise RuntimeError(
-                        "_get_slack_document_ids_and_channels: Stop signal detected"
-                    )
-
-                callback.progress("_get_slack_document_ids_and_channels", 1)
-
             if doc_metadata.perm_sync_data is None:
                 continue
             channel_id = doc_metadata.perm_sync_data["channel_id"]
@@ -123,7 +114,7 @@ def _fetch_channel_permissions(
 
 
 def slack_doc_sync(
-    cc_pair: ConnectorCredentialPair, callback: IndexingHeartbeatInterface | None
+    cc_pair: ConnectorCredentialPair,
 ) -> list[DocExternalAccess]:
     """
     Adds the external permissions to the documents in postgres
@@ -136,7 +127,7 @@ def slack_doc_sync(
     )
     user_id_to_email_map = fetch_user_id_to_email_map(slack_client)
     channel_doc_map = _get_slack_document_ids_and_channels(
-        cc_pair=cc_pair, callback=callback
+        cc_pair=cc_pair,
     )
     workspace_permissions = _fetch_workspace_permissions(
         user_id_to_email_map=user_id_to_email_map,

backend/ee/onyx/external_permissions/sync_params.py

@@ -15,13 +15,11 @@ from ee.onyx.external_permissions.slack.doc_sync import slack_doc_sync
 from onyx.access.models import DocExternalAccess
 from onyx.configs.constants import DocumentSource
 from onyx.db.models import ConnectorCredentialPair
-from onyx.indexing.indexing_heartbeat import IndexingHeartbeatInterface
 
 # Defining the input/output types for the sync functions
 DocSyncFuncType = Callable[
     [
         ConnectorCredentialPair,
-        IndexingHeartbeatInterface | None,
     ],
     list[DocExternalAccess],
 ]

backend/onyx/background/celery/apps/app_base.py

@@ -198,8 +198,7 @@ def on_celeryd_init(sender: str, conf: Any = None, **kwargs: Any) -> None:
 
 def wait_for_redis(sender: Any, **kwargs: Any) -> None:
     """Waits for redis to become ready subject to a hardcoded timeout.
-    Will raise WorkerShutdown to kill the celery worker if the timeout
-    is reached."""
+    Will raise WorkerShutdown to kill the celery worker if the timeout is reached."""
 
     r = get_redis_client(tenant_id=None)
 

backend/onyx/background/celery/celery_redis.py

@@ -91,28 +91,6 @@ def celery_find_task(task_id: str, queue: str, r: Redis) -> int:
     return False
 
 
-def celery_get_queued_task_ids(queue: str, r: Redis) -> set[str]:
-    """This is a redis specific way to build a list of tasks in a queue.
-
-    This helps us read the queue once and then efficiently look for missing tasks
-    in the queue.
-    """
-
-    task_set: set[str] = set()
-
-    for priority in range(len(OnyxCeleryPriority)):
-        queue_name = f"{queue}{CELERY_SEPARATOR}{priority}" if priority > 0 else queue
-
-        tasks = cast(list[bytes], r.lrange(queue_name, 0, -1))
-        for task in tasks:
-            task_dict: dict[str, Any] = json.loads(task.decode("utf-8"))
-            task_id = task_dict.get("headers", {}).get("id")
-            if task_id:
-                task_set.add(task_id)
-
-    return task_set
-
-
 def celery_inspect_get_workers(name_filter: str | None, app: Celery) -> list[str]:
     """Returns a list of current workers containing name_filter, or all workers if
     name_filter is None.

backend/onyx/background/celery/tasks/doc_permission_syncing/tasks.py

@@ -3,16 +3,13 @@ from datetime import datetime
 from datetime import timedelta
 from datetime import timezone
 from time import sleep
-from typing import cast
 from uuid import uuid4
 
 from celery import Celery
 from celery import shared_task
 from celery import Task
 from celery.exceptions import SoftTimeLimitExceeded
-from pydantic import ValidationError
 from redis import Redis
-from redis.exceptions import LockError
 from redis.lock import Lock as RedisLock
 from sqlalchemy.orm import Session
 
@@ -25,10 +22,6 @@ from ee.onyx.external_permissions.sync_params import (
 )
 from onyx.access.models import DocExternalAccess
 from onyx.background.celery.apps.app_base import task_logger
-from onyx.background.celery.celery_redis import celery_find_task
-from onyx.background.celery.celery_redis import celery_get_queue_length
-from onyx.background.celery.celery_redis import celery_get_queued_task_ids
-from onyx.background.celery.celery_redis import celery_get_unacked_task_ids
 from onyx.configs.app_configs import JOB_TIMEOUT
 from onyx.configs.constants import CELERY_GENERIC_BEAT_LOCK_TIMEOUT
 from onyx.configs.constants import CELERY_PERMISSIONS_SYNC_LOCK_TIMEOUT
@@ -39,7 +32,6 @@ from onyx.configs.constants import OnyxCeleryPriority
 from onyx.configs.constants import OnyxCeleryQueues
 from onyx.configs.constants import OnyxCeleryTask
 from onyx.configs.constants import OnyxRedisLocks
-from onyx.configs.constants import OnyxRedisSignals
 from onyx.db.connector import mark_cc_pair_as_permissions_synced
 from onyx.db.connector_credential_pair import get_connector_credential_pair_from_id
 from onyx.db.document import upsert_document_by_connector_credential_pair
@@ -52,19 +44,14 @@ from onyx.db.models import ConnectorCredentialPair
 from onyx.db.sync_record import insert_sync_record
 from onyx.db.sync_record import update_sync_record_status
 from onyx.db.users import batch_add_ext_perm_user_if_not_exists
-from onyx.indexing.indexing_heartbeat import IndexingHeartbeatInterface
 from onyx.redis.redis_connector import RedisConnector
-from onyx.redis.redis_connector_doc_perm_sync import RedisConnectorPermissionSync
-from onyx.redis.redis_connector_doc_perm_sync import RedisConnectorPermissionSyncPayload
+from onyx.redis.redis_connector_doc_perm_sync import (
+    RedisConnectorPermissionSyncPayload,
+)
 from onyx.redis.redis_pool import get_redis_client
-from onyx.redis.redis_pool import redis_lock_dump
-from onyx.redis.redis_pool import SCAN_ITER_COUNT_DEFAULT
-from onyx.server.utils import make_short_id
 from onyx.utils.logger import doc_permission_sync_ctx
-from onyx.utils.logger import LoggerContextVars
 from onyx.utils.logger import setup_logger
 
-
 logger = setup_logger()
 
 
@@ -118,12 +105,7 @@ def _is_external_doc_permissions_sync_due(cc_pair: ConnectorCredentialPair) -> b
     bind=True,
 )
 def check_for_doc_permissions_sync(self: Task, *, tenant_id: str | None) -> bool | None:
-    # TODO(rkuo): merge into check function after lookup table for fences is added
-
-    # we need to use celery's redis client to access its redis data
-    # (which lives on a different db number)
     r = get_redis_client(tenant_id=tenant_id)
-    r_celery: Redis = self.app.broker_connection().channel().client  # type: ignore
 
     lock_beat: RedisLock = r.lock(
         OnyxRedisLocks.CHECK_CONNECTOR_DOC_PERMISSIONS_SYNC_BEAT_LOCK,
@@ -144,32 +126,14 @@ def check_for_doc_permissions_sync(self: Task, *, tenant_id: str | None) -> bool
                 if _is_external_doc_permissions_sync_due(cc_pair):
                     cc_pair_ids_to_sync.append(cc_pair.id)
 
-        lock_beat.reacquire()
         for cc_pair_id in cc_pair_ids_to_sync:
-            payload_id = try_creating_permissions_sync_task(
+            tasks_created = try_creating_permissions_sync_task(
                 self.app, cc_pair_id, r, tenant_id
             )
-            if not payload_id:
+            if not tasks_created:
                 continue
 
-            task_logger.info(
-                f"Permissions sync queued: cc_pair={cc_pair_id} id={payload_id}"
-            )
-
-        # we want to run this less frequently than the overall task
-        lock_beat.reacquire()
-        if not r.exists(OnyxRedisSignals.VALIDATE_PERMISSION_SYNC_FENCES):
-            # clear any permission fences that don't have associated celery tasks in progress
-            # tasks can be in the queue in redis, in reserved tasks (prefetched by the worker),
-            # or be currently executing
-            try:
-                validate_permission_sync_fences(tenant_id, r, r_celery, lock_beat)
-            except Exception:
-                task_logger.exception(
-                    "Exception while validating permission sync fences"
-                )
-
-            r.set(OnyxRedisSignals.VALIDATE_PERMISSION_SYNC_FENCES, 1, ex=60)
+            task_logger.info(f"Doc permissions sync queued: cc_pair={cc_pair_id}")
     except SoftTimeLimitExceeded:
         task_logger.info(
             "Soft time limit exceeded, task is being terminated gracefully."
@@ -188,15 +152,13 @@ def try_creating_permissions_sync_task(
     cc_pair_id: int,
     r: Redis,
     tenant_id: str | None,
-) -> str | None:
-    """Returns a randomized payload id on success.
+) -> int | None:
+    """Returns an int if syncing is needed. The int represents the number of sync tasks generated.
     Returns None if no syncing is required."""
-    LOCK_TIMEOUT = 30
-
-    payload_id: str | None = None
-
     redis_connector = RedisConnector(tenant_id, cc_pair_id)
 
+    LOCK_TIMEOUT = 30
+
     lock: RedisLock = r.lock(
         DANSWER_REDIS_FUNCTION_LOCK_PREFIX + "try_generate_permissions_sync_tasks",
         timeout=LOCK_TIMEOUT,
@@ -231,13 +193,7 @@ def try_creating_permissions_sync_task(
             )
 
         # set a basic fence to start
-        redis_connector.permissions.set_active()
-        payload = RedisConnectorPermissionSyncPayload(
-            id=make_short_id(),
-            submitted=datetime.now(timezone.utc),
-            started=None,
-            celery_task_id=None,
-        )
+        payload = RedisConnectorPermissionSyncPayload(started=None, celery_task_id=None)
         redis_connector.permissions.set_fence(payload)
 
         result = app.send_task(
@@ -252,11 +208,8 @@ def try_creating_permissions_sync_task(
         )
 
         # fill in the celery task id
-        redis_connector.permissions.set_active()
         payload.celery_task_id = result.id
         redis_connector.permissions.set_fence(payload)
-
-        payload_id = payload.celery_task_id
     except Exception:
         task_logger.exception(f"Unexpected exception: cc_pair={cc_pair_id}")
         return None
@@ -264,7 +217,7 @@ def try_creating_permissions_sync_task(
         if lock.owned():
             lock.release()
 
-    return payload_id
+    return 1
 
 
 @shared_task(
@@ -285,8 +238,6 @@ def connector_permission_sync_generator_task(
     This task assumes that the task has already been properly fenced
     """
 
-    LoggerContextVars.reset()
-
     doc_permission_sync_ctx_dict = doc_permission_sync_ctx.get()
     doc_permission_sync_ctx_dict["cc_pair_id"] = cc_pair_id
     doc_permission_sync_ctx_dict["request_id"] = self.request.id
@@ -374,17 +325,12 @@ def connector_permission_sync_generator_task(
                 raise ValueError(f"No fence payload found: cc_pair={cc_pair_id}")
 
             new_payload = RedisConnectorPermissionSyncPayload(
-                id=payload.id,
-                submitted=payload.submitted,
                 started=datetime.now(timezone.utc),
                 celery_task_id=payload.celery_task_id,
             )
             redis_connector.permissions.set_fence(new_payload)
 
-            callback = PermissionSyncCallback(redis_connector, lock, r)
-            document_external_accesses: list[DocExternalAccess] = doc_sync_func(
-                cc_pair, callback
-            )
+            document_external_accesses: list[DocExternalAccess] = doc_sync_func(cc_pair)
 
             task_logger.info(
                 f"RedisConnector.permissions.generate_tasks starting. cc_pair={cc_pair_id}"
@@ -434,8 +380,6 @@ def update_external_document_permissions_task(
     connector_id: int,
     credential_id: int,
 ) -> bool:
-    start = time.monotonic()
-
     document_external_access = DocExternalAccess.from_dict(
         serialized_doc_external_access
     )
@@ -465,268 +409,16 @@ def update_external_document_permissions_task(
                     document_ids=[doc_id],
                 )
 
-            elapsed = time.monotonic() - start
-            task_logger.info(
-                f"connector_id={connector_id} "
-                f"doc={doc_id} "
-                f"action=update_permissions "
-                f"elapsed={elapsed:.2f}"
+            logger.debug(
+                f"Successfully synced postgres document permissions for {doc_id}"
             )
+        return True
     except Exception:
-        task_logger.exception(
-            f"Exception in update_external_document_permissions_task: "
-            f"connector_id={connector_id} "
-            f"doc_id={doc_id}"
+        logger.exception(
+            f"Error Syncing Document Permissions: connector_id={connector_id} doc_id={doc_id}"
         )
         return False
 
-    return True
-
-
-def validate_permission_sync_fences(
-    tenant_id: str | None,
-    r: Redis,
-    r_celery: Redis,
-    lock_beat: RedisLock,
-) -> None:
-    # building lookup table can be expensive, so we won't bother
-    # validating until the queue is small
-    PERMISSION_SYNC_VALIDATION_MAX_QUEUE_LEN = 1024
-
-    queue_len = celery_get_queue_length(
-        OnyxCeleryQueues.DOC_PERMISSIONS_UPSERT, r_celery
-    )
-    if queue_len > PERMISSION_SYNC_VALIDATION_MAX_QUEUE_LEN:
-        return
-
-    queued_upsert_tasks = celery_get_queued_task_ids(
-        OnyxCeleryQueues.DOC_PERMISSIONS_UPSERT, r_celery
-    )
-    reserved_generator_tasks = celery_get_unacked_task_ids(
-        OnyxCeleryQueues.CONNECTOR_DOC_PERMISSIONS_SYNC, r_celery
-    )
-
-    # validate all existing indexing jobs
-    for key_bytes in r.scan_iter(
-        RedisConnectorPermissionSync.FENCE_PREFIX + "*",
-        count=SCAN_ITER_COUNT_DEFAULT,
-    ):
-        lock_beat.reacquire()
-        validate_permission_sync_fence(
-            tenant_id,
-            key_bytes,
-            queued_upsert_tasks,
-            reserved_generator_tasks,
-            r,
-            r_celery,
-        )
-    return
-
-
-def validate_permission_sync_fence(
-    tenant_id: str | None,
-    key_bytes: bytes,
-    queued_tasks: set[str],
-    reserved_tasks: set[str],
-    r: Redis,
-    r_celery: Redis,
-) -> None:
-    """Checks for the error condition where an indexing fence is set but the associated celery tasks don't exist.
-    This can happen if the indexing worker hard crashes or is terminated.
-    Being in this bad state means the fence will never clear without help, so this function
-    gives the help.
-
-    How this works:
-    1. This function renews the active signal with a 5 minute TTL under the following conditions
-    1.2. When the task is seen in the redis queue
-    1.3. When the task is seen in the reserved / prefetched list
-
-    2. Externally, the active signal is renewed when:
-    2.1. The fence is created
-    2.2. The indexing watchdog checks the spawned task.
-
-    3. The TTL allows us to get through the transitions on fence startup
-    and when the task starts executing.
-
-    More TTL clarification: it is seemingly impossible to exactly query Celery for
-    whether a task is in the queue or currently executing.
-    1. An unknown task id is always returned as state PENDING.
-    2. Redis can be inspected for the task id, but the task id is gone between the time a worker receives the task
-    and the time it actually starts on the worker.
-
-    queued_tasks: the celery queue of lightweight permission sync tasks
-    reserved_tasks: prefetched tasks for sync task generator
-    """
-    # if the fence doesn't exist, there's nothing to do
-    fence_key = key_bytes.decode("utf-8")
-    cc_pair_id_str = RedisConnector.get_id_from_fence_key(fence_key)
-    if cc_pair_id_str is None:
-        task_logger.warning(
-            f"validate_permission_sync_fence - could not parse id from {fence_key}"
-        )
-        return
-
-    cc_pair_id = int(cc_pair_id_str)
-    # parse out metadata and initialize the helper class with it
-    redis_connector = RedisConnector(tenant_id, int(cc_pair_id))
-
-    # check to see if the fence/payload exists
-    if not redis_connector.permissions.fenced:
-        return
-
-    # in the cloud, the payload format may have changed ...
-    # it's a little sloppy, but just reset the fence for now if that happens
-    # TODO: add intentional cleanup/abort logic
-    try:
-        payload = redis_connector.permissions.payload
-    except ValidationError:
-        task_logger.exception(
-            "validate_permission_sync_fence - "
-            "Resetting fence because fence schema is out of date: "
-            f"cc_pair={cc_pair_id} "
-            f"fence={fence_key}"
-        )
-
-        redis_connector.permissions.reset()
-        return
-
-    if not payload:
-        return
-
-    if not payload.celery_task_id:
-        return
-
-    # OK, there's actually something for us to validate
-
-    # either the generator task must be in flight or its subtasks must be
-    found = celery_find_task(
-        payload.celery_task_id,
-        OnyxCeleryQueues.CONNECTOR_DOC_PERMISSIONS_SYNC,
-        r_celery,
-    )
-    if found:
-        # the celery task exists in the redis queue
-        redis_connector.permissions.set_active()
-        return
-
-    if payload.celery_task_id in reserved_tasks:
-        # the celery task was prefetched and is reserved within a worker
-        redis_connector.permissions.set_active()
-        return
-
-    # look up every task in the current taskset in the celery queue
-    # every entry in the taskset should have an associated entry in the celery task queue
-    # because we get the celery tasks first, the entries in our own permissions taskset
-    # should be roughly a subset of the tasks in celery
-
-    # this check isn't very exact, but should be sufficient over a period of time
-    # A single successful check over some number of attempts is sufficient.
-
-    # TODO: if the number of tasks in celery is much lower than than the taskset length
-    # we might be able to shortcut the lookup since by definition some of the tasks
-    # must not exist in celery.
-
-    tasks_scanned = 0
-    tasks_not_in_celery = 0  # a non-zero number after completing our check is bad
-
-    for member in r.sscan_iter(redis_connector.permissions.taskset_key):
-        tasks_scanned += 1
-
-        member_bytes = cast(bytes, member)
-        member_str = member_bytes.decode("utf-8")
-        if member_str in queued_tasks:
-            continue
-
-        if member_str in reserved_tasks:
-            continue
-
-        tasks_not_in_celery += 1
-
-    task_logger.info(
-        "validate_permission_sync_fence task check: "
-        f"tasks_scanned={tasks_scanned} tasks_not_in_celery={tasks_not_in_celery}"
-    )
-
-    if tasks_not_in_celery == 0:
-        redis_connector.permissions.set_active()
-        return
-
-    # we may want to enable this check if using the active task list somehow isn't good enough
-    # if redis_connector_index.generator_locked():
-    #     logger.info(f"{payload.celery_task_id} is currently executing.")
-
-    # if we get here, we didn't find any direct indication that the associated celery tasks exist,
-    # but they still might be there due to gaps in our ability to check states during transitions
-    # Checking the active signal safeguards us against these transition periods
-    # (which has a duration that allows us to bridge those gaps)
-    if redis_connector.permissions.active():
-        return
-
-    # celery tasks don't exist and the active signal has expired, possibly due to a crash. Clean it up.
-    task_logger.warning(
-        "validate_permission_sync_fence - "
-        "Resetting fence because no associated celery tasks were found: "
-        f"cc_pair={cc_pair_id} "
-        f"fence={fence_key}"
-    )
-
-    redis_connector.permissions.reset()
-    return
-
-
-class PermissionSyncCallback(IndexingHeartbeatInterface):
-    PARENT_CHECK_INTERVAL = 60
-
-    def __init__(
-        self,
-        redis_connector: RedisConnector,
-        redis_lock: RedisLock,
-        redis_client: Redis,
-    ):
-        super().__init__()
-        self.redis_connector: RedisConnector = redis_connector
-        self.redis_lock: RedisLock = redis_lock
-        self.redis_client = redis_client
-
-        self.started: datetime = datetime.now(timezone.utc)
-        self.redis_lock.reacquire()
-
-        self.last_tag: str = "PermissionSyncCallback.__init__"
-        self.last_lock_reacquire: datetime = datetime.now(timezone.utc)
-        self.last_lock_monotonic = time.monotonic()
-
-    def should_stop(self) -> bool:
-        if self.redis_connector.stop.fenced:
-            return True
-
-        return False
-
-    def progress(self, tag: str, amount: int) -> None:
-        try:
-            self.redis_connector.permissions.set_active()
-
-            current_time = time.monotonic()
-            if current_time - self.last_lock_monotonic >= (
-                CELERY_GENERIC_BEAT_LOCK_TIMEOUT / 4
-            ):
-                self.redis_lock.reacquire()
-                self.last_lock_reacquire = datetime.now(timezone.utc)
-                self.last_lock_monotonic = time.monotonic()
-
-            self.last_tag = tag
-        except LockError:
-            logger.exception(
-                f"PermissionSyncCallback - lock.reacquire exceptioned: "
-                f"lock_timeout={self.redis_lock.timeout} "
-                f"start={self.started} "
-                f"last_tag={self.last_tag} "
-                f"last_reacquired={self.last_lock_reacquire} "
-                f"now={datetime.now(timezone.utc)}"
-            )
-
-            redis_lock_dump(self.redis_lock, self.redis_client)
-            raise
-
 
 """Monitoring CCPair permissions utils, called in monitor_vespa_sync"""
 
@@ -752,36 +444,20 @@ def monitor_ccpair_permissions_taskset(
     if initial is None:
         return
 
-    try:
-        payload = redis_connector.permissions.payload
-    except ValidationError:
-        task_logger.exception(
-            "Permissions sync payload failed to validate. "
-            "Schema may have been updated."
-        )
-        return
-
-    if not payload:
-        return
-
     remaining = redis_connector.permissions.get_remaining()
     task_logger.info(
-        f"Permissions sync progress: "
-        f"cc_pair={cc_pair_id} "
-        f"id={payload.id} "
-        f"remaining={remaining} "
-        f"initial={initial}"
+        f"Permissions sync progress: cc_pair={cc_pair_id} remaining={remaining} initial={initial}"
     )
     if remaining > 0:
         return
 
-    mark_cc_pair_as_permissions_synced(db_session, int(cc_pair_id), payload.started)
-    task_logger.info(
-        f"Permissions sync finished: "
-        f"cc_pair={cc_pair_id} "
-        f"id={payload.id} "
-        f"num_synced={initial}"
+    payload: RedisConnectorPermissionSyncPayload | None = (
+        redis_connector.permissions.payload
     )
+    start_time: datetime | None = payload.started if payload else None
+
+    mark_cc_pair_as_permissions_synced(db_session, int(cc_pair_id), start_time)
+    task_logger.info(f"Successfully synced permissions for cc_pair={cc_pair_id}")
 
     update_sync_record_status(
         db_session=db_session,

backend/onyx/background/celery/tasks/external_group_syncing/tasks.py

@@ -1,4 +1,3 @@
-import time
 from datetime import datetime
 from datetime import timedelta
 from datetime import timezone
@@ -10,7 +9,6 @@ from celery import Task
 from celery.exceptions import SoftTimeLimitExceeded
 from redis import Redis
 from redis.lock import Lock as RedisLock
-from sqlalchemy.orm import Session
 
 from ee.onyx.db.connector_credential_pair import get_all_auto_sync_cc_pairs
 from ee.onyx.db.connector_credential_pair import get_cc_pairs_by_source
@@ -22,12 +20,9 @@ from ee.onyx.external_permissions.sync_params import (
     GROUP_PERMISSIONS_IS_CC_PAIR_AGNOSTIC,
 )
 from onyx.background.celery.apps.app_base import task_logger
-from onyx.background.celery.celery_redis import celery_find_task
-from onyx.background.celery.celery_redis import celery_get_unacked_task_ids
 from onyx.configs.app_configs import JOB_TIMEOUT
 from onyx.configs.constants import CELERY_EXTERNAL_GROUP_SYNC_LOCK_TIMEOUT
 from onyx.configs.constants import CELERY_GENERIC_BEAT_LOCK_TIMEOUT
-from onyx.configs.constants import CELERY_TASK_WAIT_FOR_FENCE_TIMEOUT
 from onyx.configs.constants import DANSWER_REDIS_FUNCTION_LOCK_PREFIX
 from onyx.configs.constants import OnyxCeleryPriority
 from onyx.configs.constants import OnyxCeleryQueues
@@ -44,12 +39,10 @@ from onyx.db.models import ConnectorCredentialPair
 from onyx.db.sync_record import insert_sync_record
 from onyx.db.sync_record import update_sync_record_status
 from onyx.redis.redis_connector import RedisConnector
-from onyx.redis.redis_connector_ext_group_sync import RedisConnectorExternalGroupSync
 from onyx.redis.redis_connector_ext_group_sync import (
     RedisConnectorExternalGroupSyncPayload,
 )
 from onyx.redis.redis_pool import get_redis_client
-from onyx.redis.redis_pool import SCAN_ITER_COUNT_DEFAULT
 from onyx.utils.logger import setup_logger
 
 logger = setup_logger()
@@ -109,10 +102,6 @@ def _is_external_group_sync_due(cc_pair: ConnectorCredentialPair) -> bool:
 def check_for_external_group_sync(self: Task, *, tenant_id: str | None) -> bool | None:
     r = get_redis_client(tenant_id=tenant_id)
 
-    # we need to use celery's redis client to access its redis data
-    # (which lives on a different db number)
-    # r_celery: Redis = self.app.broker_connection().channel().client  # type: ignore
-
     lock_beat: RedisLock = r.lock(
         OnyxRedisLocks.CHECK_CONNECTOR_EXTERNAL_GROUP_SYNC_BEAT_LOCK,
         timeout=CELERY_GENERIC_BEAT_LOCK_TIMEOUT,
@@ -147,7 +136,6 @@ def check_for_external_group_sync(self: Task, *, tenant_id: str | None) -> bool
                 if _is_external_group_sync_due(cc_pair):
                     cc_pair_ids_to_sync.append(cc_pair.id)
 
-        lock_beat.reacquire()
         for cc_pair_id in cc_pair_ids_to_sync:
             tasks_created = try_creating_external_group_sync_task(
                 self.app, cc_pair_id, r, tenant_id
@@ -156,23 +144,6 @@ def check_for_external_group_sync(self: Task, *, tenant_id: str | None) -> bool
                 continue
 
             task_logger.info(f"External group sync queued: cc_pair={cc_pair_id}")
-
-        # we want to run this less frequently than the overall task
-        # lock_beat.reacquire()
-        # if not r.exists(OnyxRedisSignals.VALIDATE_EXTERNAL_GROUP_SYNC_FENCES):
-        #     # clear any indexing fences that don't have associated celery tasks in progress
-        #     # tasks can be in the queue in redis, in reserved tasks (prefetched by the worker),
-        #     # or be currently executing
-        #     try:
-        #         validate_external_group_sync_fences(
-        #             tenant_id, self.app, r, r_celery, lock_beat
-        #         )
-        #     except Exception:
-        #         task_logger.exception(
-        #             "Exception while validating external group sync fences"
-        #         )
-
-        #     r.set(OnyxRedisSignals.VALIDATE_EXTERNAL_GROUP_SYNC_FENCES, 1, ex=60)
     except SoftTimeLimitExceeded:
         task_logger.info(
             "Soft time limit exceeded, task is being terminated gracefully."
@@ -215,12 +186,6 @@ def try_creating_external_group_sync_task(
         redis_connector.external_group_sync.generator_clear()
         redis_connector.external_group_sync.taskset_clear()
 
-        payload = RedisConnectorExternalGroupSyncPayload(
-            submitted=datetime.now(timezone.utc),
-            started=None,
-            celery_task_id=None,
-        )
-
         custom_task_id = f"{redis_connector.external_group_sync.taskset_key}_{uuid4()}"
 
         result = app.send_task(
@@ -234,6 +199,11 @@ def try_creating_external_group_sync_task(
             priority=OnyxCeleryPriority.HIGH,
         )
 
+        payload = RedisConnectorExternalGroupSyncPayload(
+            started=datetime.now(timezone.utc),
+            celery_task_id=result.id,
+        )
+
         # create before setting fence to avoid race condition where the monitoring
         # task updates the sync record before it is created
         with get_session_with_tenant(tenant_id) as db_session:
@@ -243,8 +213,8 @@ def try_creating_external_group_sync_task(
                 sync_type=SyncType.EXTERNAL_GROUP,
             )
 
-        payload.celery_task_id = result.id
         redis_connector.external_group_sync.set_fence(payload)
+
     except Exception:
         task_logger.exception(
             f"Unexpected exception while trying to create external group sync task: cc_pair={cc_pair_id}"
@@ -271,7 +241,7 @@ def connector_external_group_sync_generator_task(
     tenant_id: str | None,
 ) -> None:
     """
-    External group sync task for a given connector credential pair
+    Permission sync task that handles external group syncing for a given connector credential pair
     This task assumes that the task has already been properly fenced
     """
 
@@ -279,59 +249,19 @@ def connector_external_group_sync_generator_task(
 
     r = get_redis_client(tenant_id=tenant_id)
 
-    # this wait is needed to avoid a race condition where
-    # the primary worker sends the task and it is immediately executed
-    # before the primary worker can finalize the fence
-    start = time.monotonic()
-    while True:
-        if time.monotonic() - start > CELERY_TASK_WAIT_FOR_FENCE_TIMEOUT:
-            raise ValueError(
-                f"connector_external_group_sync_generator_task - timed out waiting for fence to be ready: "
-                f"fence={redis_connector.external_group_sync.fence_key}"
-            )
-
-        if not redis_connector.external_group_sync.fenced:  # The fence must exist
-            raise ValueError(
-                f"connector_external_group_sync_generator_task - fence not found: "
-                f"fence={redis_connector.external_group_sync.fence_key}"
-            )
-
-        payload = redis_connector.external_group_sync.payload  # The payload must exist
-        if not payload:
-            raise ValueError(
-                "connector_external_group_sync_generator_task: payload invalid or not found"
-            )
-
-        if payload.celery_task_id is None:
-            logger.info(
-                f"connector_external_group_sync_generator_task - Waiting for fence: "
-                f"fence={redis_connector.external_group_sync.fence_key}"
-            )
-            time.sleep(1)
-            continue
-
-        logger.info(
-            f"connector_external_group_sync_generator_task - Fence found, continuing...: "
-            f"fence={redis_connector.external_group_sync.fence_key}"
-        )
-        break
-
     lock: RedisLock = r.lock(
         OnyxRedisLocks.CONNECTOR_EXTERNAL_GROUP_SYNC_LOCK_PREFIX
         + f"_{redis_connector.id}",
         timeout=CELERY_EXTERNAL_GROUP_SYNC_LOCK_TIMEOUT,
     )
 
-    acquired = lock.acquire(blocking=False)
-    if not acquired:
-        task_logger.warning(
-            f"External group sync task already running, exiting...: cc_pair={cc_pair_id}"
-        )
-        return None
-
     try:
-        payload.started = datetime.now(timezone.utc)
-        redis_connector.external_group_sync.set_fence(payload)
+        acquired = lock.acquire(blocking=False)
+        if not acquired:
+            task_logger.warning(
+                f"External group sync task already running, exiting...: cc_pair={cc_pair_id}"
+            )
+            return None
 
         with get_session_with_tenant(tenant_id) as db_session:
             cc_pair = get_connector_credential_pair_from_id(
@@ -400,135 +330,3 @@ def connector_external_group_sync_generator_task(
         redis_connector.external_group_sync.set_fence(None)
         if lock.owned():
             lock.release()
-
-
-def validate_external_group_sync_fences(
-    tenant_id: str | None,
-    celery_app: Celery,
-    r: Redis,
-    r_celery: Redis,
-    lock_beat: RedisLock,
-) -> None:
-    reserved_sync_tasks = celery_get_unacked_task_ids(
-        OnyxCeleryQueues.CONNECTOR_EXTERNAL_GROUP_SYNC, r_celery
-    )
-
-    # validate all existing indexing jobs
-    for key_bytes in r.scan_iter(
-        RedisConnectorExternalGroupSync.FENCE_PREFIX + "*",
-        count=SCAN_ITER_COUNT_DEFAULT,
-    ):
-        lock_beat.reacquire()
-        with get_session_with_tenant(tenant_id) as db_session:
-            validate_external_group_sync_fence(
-                tenant_id,
-                key_bytes,
-                reserved_sync_tasks,
-                r_celery,
-                db_session,
-            )
-    return
-
-
-def validate_external_group_sync_fence(
-    tenant_id: str | None,
-    key_bytes: bytes,
-    reserved_tasks: set[str],
-    r_celery: Redis,
-    db_session: Session,
-) -> None:
-    """Checks for the error condition where an indexing fence is set but the associated celery tasks don't exist.
-    This can happen if the indexing worker hard crashes or is terminated.
-    Being in this bad state means the fence will never clear without help, so this function
-    gives the help.
-
-    How this works:
-    1. This function renews the active signal with a 5 minute TTL under the following conditions
-    1.2. When the task is seen in the redis queue
-    1.3. When the task is seen in the reserved / prefetched list
-
-    2. Externally, the active signal is renewed when:
-    2.1. The fence is created
-    2.2. The indexing watchdog checks the spawned task.
-
-    3. The TTL allows us to get through the transitions on fence startup
-    and when the task starts executing.
-
-    More TTL clarification: it is seemingly impossible to exactly query Celery for
-    whether a task is in the queue or currently executing.
-    1. An unknown task id is always returned as state PENDING.
-    2. Redis can be inspected for the task id, but the task id is gone between the time a worker receives the task
-    and the time it actually starts on the worker.
-    """
-    # if the fence doesn't exist, there's nothing to do
-    fence_key = key_bytes.decode("utf-8")
-    cc_pair_id_str = RedisConnector.get_id_from_fence_key(fence_key)
-    if cc_pair_id_str is None:
-        task_logger.warning(
-            f"validate_external_group_sync_fence - could not parse id from {fence_key}"
-        )
-        return
-
-    cc_pair_id = int(cc_pair_id_str)
-
-    # parse out metadata and initialize the helper class with it
-    redis_connector = RedisConnector(tenant_id, int(cc_pair_id))
-
-    # check to see if the fence/payload exists
-    if not redis_connector.external_group_sync.fenced:
-        return
-
-    payload = redis_connector.external_group_sync.payload
-    if not payload:
-        return
-
-    # OK, there's actually something for us to validate
-
-    if payload.celery_task_id is None:
-        # the fence is just barely set up.
-        # if redis_connector_index.active():
-        #     return
-
-        # it would be odd to get here as there isn't that much that can go wrong during
-        # initial fence setup, but it's still worth making sure we can recover
-        logger.info(
-            "validate_external_group_sync_fence - "
-            f"Resetting fence in basic state without any activity: fence={fence_key}"
-        )
-        redis_connector.external_group_sync.reset()
-        return
-
-    found = celery_find_task(
-        payload.celery_task_id, OnyxCeleryQueues.CONNECTOR_EXTERNAL_GROUP_SYNC, r_celery
-    )
-    if found:
-        # the celery task exists in the redis queue
-        # redis_connector_index.set_active()
-        return
-
-    if payload.celery_task_id in reserved_tasks:
-        # the celery task was prefetched and is reserved within the indexing worker
-        # redis_connector_index.set_active()
-        return
-
-    # we may want to enable this check if using the active task list somehow isn't good enough
-    # if redis_connector_index.generator_locked():
-    #     logger.info(f"{payload.celery_task_id} is currently executing.")
-
-    # if we get here, we didn't find any direct indication that the associated celery tasks exist,
-    # but they still might be there due to gaps in our ability to check states during transitions
-    # Checking the active signal safeguards us against these transition periods
-    # (which has a duration that allows us to bridge those gaps)
-    # if redis_connector_index.active():
-    # return
-
-    # celery tasks don't exist and the active signal has expired, possibly due to a crash. Clean it up.
-    logger.warning(
-        "validate_external_group_sync_fence - "
-        "Resetting fence because no associated celery tasks were found: "
-        f"cc_pair={cc_pair_id} "
-        f"fence={fence_key}"
-    )
-
-    redis_connector.external_group_sync.reset()
-    return

backend/onyx/background/celery/tasks/monitoring/tasks.py

@@ -39,7 +39,6 @@ from onyx.redis.redis_pool import get_redis_client
 from onyx.redis.redis_pool import redis_lock_dump
 from onyx.utils.telemetry import optional_telemetry
 from onyx.utils.telemetry import RecordType
-from shared_configs.contextvars import CURRENT_TENANT_ID_CONTEXTVAR
 
 
 _MONITORING_SOFT_TIME_LIMIT = 60 * 5  # 5 minutes
@@ -658,9 +657,6 @@ def monitor_background_processes(self: Task, *, tenant_id: str | None) -> None:
     - Syncing speed metrics
     - Worker status and task counts
     """
-    if tenant_id is not None:
-        CURRENT_TENANT_ID_CONTEXTVAR.set(tenant_id)
-
     task_logger.info("Starting background monitoring")
     r = get_redis_client(tenant_id=tenant_id)
 
@@ -692,13 +688,11 @@ def monitor_background_processes(self: Task, *, tenant_id: str | None) -> None:
                 metrics = metric_fn()
                 for metric in metrics:
                     # double check to make sure we aren't double-emitting metrics
-                    if metric.key is None or not _has_metric_been_emitted(
+                    if metric.key is not None and not _has_metric_been_emitted(
                         redis_std, metric.key
                     ):
                         metric.log()
                         metric.emit(tenant_id)
-
-                    if metric.key is not None:
                         _mark_metric_as_emitted(redis_std, metric.key)
 
         task_logger.info("Successfully collected background metrics")

backend/onyx/background/celery/tasks/pruning/tasks.py

@@ -39,7 +39,6 @@ from onyx.db.sync_record import insert_sync_record
 from onyx.db.sync_record import update_sync_record_status
 from onyx.redis.redis_connector import RedisConnector
 from onyx.redis.redis_pool import get_redis_client
-from onyx.utils.logger import LoggerContextVars
 from onyx.utils.logger import pruning_ctx
 from onyx.utils.logger import setup_logger
 
@@ -252,8 +251,6 @@ def connector_pruning_generator_task(
     and compares those IDs to locally stored documents and deletes all locally stored IDs missing
     from the most recently pulled document ID list"""
 
-    LoggerContextVars.reset()
-
     pruning_ctx_dict = pruning_ctx.get()
     pruning_ctx_dict["cc_pair_id"] = cc_pair_id
     pruning_ctx_dict["request_id"] = self.request.id
@@ -402,7 +399,7 @@ def monitor_ccpair_pruning_taskset(
 
     mark_ccpair_as_pruned(int(cc_pair_id), db_session)
     task_logger.info(
-        f"Connector pruning finished: cc_pair={cc_pair_id} num_pruned={initial}"
+        f"Successfully pruned connector credential pair. cc_pair={cc_pair_id}"
     )
 
     update_sync_record_status(

backend/onyx/background/celery/tasks/shared/tasks.py

@@ -75,8 +75,6 @@ def document_by_cc_pair_cleanup_task(
     """
     task_logger.debug(f"Task start: doc={document_id}")
 
-    start = time.monotonic()
-
     try:
         with get_session_with_tenant(tenant_id) as db_session:
             action = "skip"
@@ -156,13 +154,11 @@ def document_by_cc_pair_cleanup_task(
 
             db_session.commit()
 
-            elapsed = time.monotonic() - start
             task_logger.info(
                 f"doc={document_id} "
                 f"action={action} "
                 f"refcount={count} "
-                f"chunks={chunks_affected} "
-                f"elapsed={elapsed:.2f}"
+                f"chunks={chunks_affected}"
             )
     except SoftTimeLimitExceeded:
         task_logger.info(f"SoftTimeLimitExceeded exception. doc={document_id}")

backend/onyx/background/celery/tasks/vespa/tasks.py

@@ -989,10 +989,6 @@ def monitor_vespa_sync(self: Task, tenant_id: str | None) -> bool | None:
         task_logger.info(
             "Soft time limit exceeded, task is being terminated gracefully."
         )
-        return False
-    except Exception:
-        task_logger.exception("monitor_vespa_sync exceptioned.")
-        return False
     finally:
         if lock_beat.owned():
             lock_beat.release()
@@ -1082,7 +1078,6 @@ def vespa_metadata_sync_task(
             )
     except SoftTimeLimitExceeded:
         task_logger.info(f"SoftTimeLimitExceeded exception. doc={document_id}")
-        return False
     except Exception as ex:
         if isinstance(ex, RetryError):
             task_logger.warning(

backend/onyx/configs/app_configs.py

@@ -478,12 +478,6 @@ INDEXING_SIZE_WARNING_THRESHOLD = int(
 # 0 disables this behavior and is the default.
 INDEXING_TRACER_INTERVAL = int(os.environ.get("INDEXING_TRACER_INTERVAL") or 0)
 
-# Enable multi-threaded embedding model calls for parallel processing
-# Note: only applies for API-based embedding models
-INDEXING_EMBEDDING_MODEL_NUM_THREADS = int(
-    os.environ.get("INDEXING_EMBEDDING_MODEL_NUM_THREADS") or 1
-)
-
 # During an indexing attempt, specifies the number of batches which are allowed to
 # exception without aborting the attempt.
 INDEXING_EXCEPTION_LIMIT = int(os.environ.get("INDEXING_EXCEPTION_LIMIT") or 0)
@@ -617,8 +611,3 @@ POD_NAMESPACE = os.environ.get("POD_NAMESPACE")
 DEV_MODE = os.environ.get("DEV_MODE", "").lower() == "true"
 
 TEST_ENV = os.environ.get("TEST_ENV", "").lower() == "true"
-
-# Set to true to mock LLM responses for testing purposes
-MOCK_LLM_RESPONSE = (
-    os.environ.get("MOCK_LLM_RESPONSE") if os.environ.get("MOCK_LLM_RESPONSE") else None
-)

backend/onyx/configs/constants.py

@@ -300,8 +300,6 @@ class OnyxRedisLocks:
 
 class OnyxRedisSignals:
     VALIDATE_INDEXING_FENCES = "signal:validate_indexing_fences"
-    VALIDATE_EXTERNAL_GROUP_SYNC_FENCES = "signal:validate_external_group_sync_fences"
-    VALIDATE_PERMISSION_SYNC_FENCES = "signal:validate_permission_sync_fences"
 
 
 class OnyxCeleryPriority(int, Enum):

backend/onyx/connectors/airtable/airtable_connector.py

@@ -1,7 +1,3 @@
-import contextvars
-from concurrent.futures import as_completed
-from concurrent.futures import Future
-from concurrent.futures import ThreadPoolExecutor
 from io import BytesIO
 from typing import Any
 
@@ -70,25 +66,18 @@ class AirtableConnector(LoadConnector):
         self.base_id = base_id
         self.table_name_or_id = table_name_or_id
         self.batch_size = batch_size
-        self._airtable_client: AirtableApi | None = None
+        self.airtable_client: AirtableApi | None = None
         self.treat_all_non_attachment_fields_as_metadata = (
             treat_all_non_attachment_fields_as_metadata
         )
 
     def load_credentials(self, credentials: dict[str, Any]) -> dict[str, Any] | None:
-        self._airtable_client = AirtableApi(credentials["airtable_access_token"])
+        self.airtable_client = AirtableApi(credentials["airtable_access_token"])
         return None
 
-    @property
-    def airtable_client(self) -> AirtableApi:
-        if not self._airtable_client:
-            raise AirtableClientNotSetUpError()
-        return self._airtable_client
-
+    @staticmethod
     def _extract_field_values(
-        self,
         field_id: str,
-        field_name: str,
         field_info: Any,
         field_type: str,
         base_id: str,
@@ -127,33 +116,13 @@ class AirtableConnector(LoadConnector):
                     backoff=2,
                     max_delay=10,
                 )
-                def get_attachment_with_retry(url: str, record_id: str) -> bytes | None:
-                    try:
-                        attachment_response = requests.get(url)
-                        attachment_response.raise_for_status()
+                def get_attachment_with_retry(url: str) -> bytes | None:
+                    attachment_response = requests.get(url)
+                    if attachment_response.status_code == 200:
                         return attachment_response.content
-                    except requests.exceptions.HTTPError as e:
-                        if e.response.status_code == 410:
-                            logger.info(f"Refreshing attachment for {filename}")
-                            # Re-fetch the record to get a fresh URL
-                            refreshed_record = self.airtable_client.table(
-                                base_id, table_id
-                            ).get(record_id)
-                            for refreshed_attachment in refreshed_record["fields"][
-                                field_name
-                            ]:
-                                if refreshed_attachment.get("filename") == filename:
-                                    new_url = refreshed_attachment.get("url")
-                                    if new_url:
-                                        attachment_response = requests.get(new_url)
-                                        attachment_response.raise_for_status()
-                                        return attachment_response.content
+                    return None
 
-                            logger.error(f"Failed to refresh attachment for {filename}")
-
-                        raise
-
-                attachment_content = get_attachment_with_retry(url, record_id)
+                attachment_content = get_attachment_with_retry(url)
                 if attachment_content:
                     try:
                         file_ext = get_file_ext(filename)
@@ -237,7 +206,6 @@ class AirtableConnector(LoadConnector):
         # Get the value(s) for the field
         field_value_and_links = self._extract_field_values(
             field_id=field_id,
-            field_name=field_name,
             field_info=field_info,
             field_type=field_type,
             base_id=self.base_id,
@@ -306,11 +274,6 @@ class AirtableConnector(LoadConnector):
             field_val = fields.get(field_name)
             field_type = field_schema.type
 
-            logger.debug(
-                f"Processing field '{field_name}' of type '{field_type}' "
-                f"for record '{record_id}'."
-            )
-
             field_sections, field_metadata = self._process_field(
                 field_id=field_schema.id,
                 field_name=field_name,
@@ -364,47 +327,19 @@ class AirtableConnector(LoadConnector):
                 primary_field_name = field.name
                 break
 
-        logger.info(f"Starting to process Airtable records for {table.name}.")
+        record_documents: list[Document] = []
+        for record in records:
+            document = self._process_record(
+                record=record,
+                table_schema=table_schema,
+                primary_field_name=primary_field_name,
+            )
+            if document:
+                record_documents.append(document)
 
-        # Process records in parallel batches using ThreadPoolExecutor
-        PARALLEL_BATCH_SIZE = 8
-        max_workers = min(PARALLEL_BATCH_SIZE, len(records))
+            if len(record_documents) >= self.batch_size:
+                yield record_documents
+                record_documents = []
 
-        # Process records in batches
-        for i in range(0, len(records), PARALLEL_BATCH_SIZE):
-            batch_records = records[i : i + PARALLEL_BATCH_SIZE]
-            record_documents: list[Document] = []
-
-            with ThreadPoolExecutor(max_workers=max_workers) as executor:
-                # Submit batch tasks
-                future_to_record: dict[Future, RecordDict] = {}
-                for record in batch_records:
-                    # Capture the current context so that the thread gets the current tenant ID
-                    current_context = contextvars.copy_context()
-                    future_to_record[
-                        executor.submit(
-                            current_context.run,
-                            self._process_record,
-                            record=record,
-                            table_schema=table_schema,
-                            primary_field_name=primary_field_name,
-                        )
-                    ] = record
-
-                # Wait for all tasks in this batch to complete
-                for future in as_completed(future_to_record):
-                    record = future_to_record[future]
-                    try:
-                        document = future.result()
-                        if document:
-                            record_documents.append(document)
-                    except Exception as e:
-                        logger.exception(f"Failed to process record {record['id']}")
-                        raise e
-
-            yield record_documents
-            record_documents = []
-
-        # Yield any remaining records
         if record_documents:
             yield record_documents

backend/onyx/connectors/connector_runner.py

@@ -1,5 +1,4 @@
 import sys
-import time
 from datetime import datetime
 
 from onyx.connectors.interfaces import BaseConnector
@@ -46,17 +45,7 @@ class ConnectorRunner:
     def run(self) -> GenerateDocumentsOutput:
         """Adds additional exception logging to the connector."""
         try:
-            start = time.monotonic()
-            for batch in self.doc_batch_generator:
-                # to know how long connector is taking
-                logger.debug(
-                    f"Connector took {time.monotonic() - start} seconds to build a batch."
-                )
-
-                yield batch
-
-                start = time.monotonic()
-
+            yield from self.doc_batch_generator
         except Exception:
             exc_type, _, exc_traceback = sys.exc_info()
 

backend/onyx/connectors/models.py

@@ -150,16 +150,6 @@ class Document(DocumentBase):
     id: str  # This must be unique or during indexing/reindexing, chunks will be overwritten
     source: DocumentSource
 
-    def get_total_char_length(self) -> int:
-        """Calculate the total character length of the document including sections, metadata, and identifiers."""
-        section_length = sum(len(section.text) for section in self.sections)
-        identifier_length = len(self.semantic_identifier) + len(self.title or "")
-        metadata_length = sum(
-            len(k) + len(v) if isinstance(v, str) else len(k) + sum(len(x) for x in v)
-            for k, v in self.metadata.items()
-        )
-        return section_length + identifier_length + metadata_length
-
     def to_short_descriptor(self) -> str:
         """Used when logging the identity of a document"""
         return f"ID: '{self.id}'; Semantic ID: '{self.semantic_identifier}'"

backend/onyx/db/models.py

@@ -150,7 +150,6 @@ class User(SQLAlchemyBaseUserTableUUID, Base):
 
     # if specified, controls the assistants that are shown to the user + their order
     # if not specified, all assistants are shown
-    temperature_override_enabled: Mapped[bool] = mapped_column(Boolean, default=False)
     auto_scroll: Mapped[bool] = mapped_column(Boolean, default=True)
     shortcut_enabled: Mapped[bool] = mapped_column(Boolean, default=True)
     chosen_assistants: Mapped[list[int] | None] = mapped_column(
@@ -162,7 +161,9 @@ class User(SQLAlchemyBaseUserTableUUID, Base):
     hidden_assistants: Mapped[list[int]] = mapped_column(
         postgresql.JSONB(), nullable=False, default=[]
     )
-
+    recent_assistants: Mapped[list[dict]] = mapped_column(
+        postgresql.JSONB(), nullable=False, default=list, server_default="[]"
+    )
     pinned_assistants: Mapped[list[int] | None] = mapped_column(
         postgresql.JSONB(), nullable=True, default=None
     )
@@ -1116,10 +1117,6 @@ class ChatSession(Base):
     llm_override: Mapped[LLMOverride | None] = mapped_column(
         PydanticType(LLMOverride), nullable=True
     )
-
-    # The latest temperature override specified by the user
-    temperature_override: Mapped[float | None] = mapped_column(Float, nullable=True)
-
     prompt_override: Mapped[PromptOverride | None] = mapped_column(
         PydanticType(PromptOverride), nullable=True
     )

backend/onyx/indexing/indexing_pipeline.py

@@ -380,15 +380,6 @@ def index_doc_batch(
             new_docs=0, total_docs=len(filtered_documents), total_chunks=0
         )
 
-    doc_descriptors = [
-        {
-            "doc_id": doc.id,
-            "doc_length": doc.get_total_char_length(),
-        }
-        for doc in ctx.updatable_docs
-    ]
-    logger.debug(f"Starting indexing process for documents: {doc_descriptors}")
-
     logger.debug("Starting chunking")
     chunks: list[DocAwareChunk] = chunker.chunk(ctx.updatable_docs)
 

backend/onyx/key_value_store/factory.py

@@ -2,7 +2,7 @@ from onyx.key_value_store.interface import KeyValueStore
 from onyx.key_value_store.store import PgRedisKVStore
 
 
-def get_kv_store() -> KeyValueStore:
+def get_kv_store(tenant_id: str | None = None) -> KeyValueStore:
     # In the Multi Tenant case, the tenant context is picked up automatically, it does not need to be passed in
     # It's read from the global thread level variable
-    return PgRedisKVStore()
+    return PgRedisKVStore(tenant_id=tenant_id)

backend/onyx/key_value_store/store.py

@@ -18,7 +18,7 @@ from onyx.utils.logger import setup_logger
 from onyx.utils.special_types import JSON_ro
 from shared_configs.configs import MULTI_TENANT
 from shared_configs.configs import POSTGRES_DEFAULT_SCHEMA
-from shared_configs.contextvars import get_current_tenant_id
+from shared_configs.contextvars import CURRENT_TENANT_ID_CONTEXTVAR
 
 logger = setup_logger()
 
@@ -28,8 +28,10 @@ KV_REDIS_KEY_EXPIRATION = 60 * 60 * 24  # 1 Day
 
 
 class PgRedisKVStore(KeyValueStore):
-    def __init__(self, redis_client: Redis | None = None) -> None:
-        self.tenant_id = get_current_tenant_id()
+    def __init__(
+        self, redis_client: Redis | None = None, tenant_id: str | None = None
+    ) -> None:
+        self.tenant_id = tenant_id or CURRENT_TENANT_ID_CONTEXTVAR.get()
 
         # If no redis_client is provided, fall back to the context var
         if redis_client is not None:

backend/onyx/llm/chat_llm.py

@@ -26,7 +26,6 @@ from langchain_core.messages.tool import ToolMessage
 from langchain_core.prompt_values import PromptValue
 
 from onyx.configs.app_configs import LOG_DANSWER_MODEL_INTERACTIONS
-from onyx.configs.app_configs import MOCK_LLM_RESPONSE
 from onyx.configs.model_configs import (
     DISABLE_LITELLM_STREAMING,
 )
@@ -388,7 +387,6 @@ class DefaultMultiLLM(LLM):
 
         try:
             return litellm.completion(
-                mock_response=MOCK_LLM_RESPONSE,
                 # model choice
                 model=f"{self.config.model_provider}/{self.config.deployment_name or self.config.model_name}",
                 # NOTE: have to pass in None instead of empty string for these

backend/onyx/main.py

@@ -109,9 +109,7 @@ from onyx.utils.variable_functionality import global_version
 from onyx.utils.variable_functionality import set_is_ee_based_on_env_variable
 from shared_configs.configs import CORS_ALLOWED_ORIGIN
 from shared_configs.configs import MULTI_TENANT
-from shared_configs.configs import POSTGRES_DEFAULT_SCHEMA
 from shared_configs.configs import SENTRY_DSN
-from shared_configs.contextvars import CURRENT_TENANT_ID_CONTEXTVAR
 
 logger = setup_logger()
 
@@ -214,8 +212,7 @@ async def lifespan(app: FastAPI) -> AsyncGenerator[None, None]:
 
     if not MULTI_TENANT:
         # We cache this at the beginning so there is no delay in the first telemetry
-        CURRENT_TENANT_ID_CONTEXTVAR.set(POSTGRES_DEFAULT_SCHEMA)
-        get_or_generate_uuid()
+        get_or_generate_uuid(tenant_id=None)
 
         # If we are multi-tenant, we need to only set up initial public tables
         with Session(engine) as db_session:

backend/onyx/natural_language_processing/search_nlp_models.py

@@ -1,8 +1,6 @@
 import threading
 import time
 from collections.abc import Callable
-from concurrent.futures import as_completed
-from concurrent.futures import ThreadPoolExecutor
 from functools import wraps
 from typing import Any
 
@@ -13,7 +11,6 @@ from requests import RequestException
 from requests import Response
 from retry import retry
 
-from onyx.configs.app_configs import INDEXING_EMBEDDING_MODEL_NUM_THREADS
 from onyx.configs.app_configs import LARGE_CHUNK_RATIO
 from onyx.configs.app_configs import SKIP_WARM_UP
 from onyx.configs.model_configs import BATCH_SIZE_ENCODE_CHUNKS
@@ -158,7 +155,6 @@ class EmbeddingModel:
         text_type: EmbedTextType,
         batch_size: int,
         max_seq_length: int,
-        num_threads: int = INDEXING_EMBEDDING_MODEL_NUM_THREADS,
     ) -> list[Embedding]:
         text_batches = batch_list(texts, batch_size)
 
@@ -167,14 +163,12 @@ class EmbeddingModel:
         )
 
         embeddings: list[Embedding] = []
-
-        def process_batch(
-            batch_idx: int, text_batch: list[str]
-        ) -> tuple[int, list[Embedding]]:
+        for idx, text_batch in enumerate(text_batches, start=1):
             if self.callback:
                 if self.callback.should_stop():
                     raise RuntimeError("_batch_encode_texts detected stop signal")
 
+            logger.debug(f"Encoding batch {idx} of {len(text_batches)}")
             embed_request = EmbedRequest(
                 model_name=self.model_name,
                 texts=text_batch,
@@ -190,52 +184,11 @@ class EmbeddingModel:
                 api_url=self.api_url,
             )
 
-            start_time = time.time()
             response = self._make_model_server_request(embed_request)
-            end_time = time.time()
-
-            processing_time = end_time - start_time
-            logger.info(
-                f"Batch {batch_idx} processing time: {processing_time:.2f} seconds"
-            )
-
-            return batch_idx, response.embeddings
-
-        # only multi thread if:
-        #   1. num_threads is greater than 1
-        #   2. we are using an API-based embedding model (provider_type is not None)
-        #   3. there are more than 1 batch (no point in threading if only 1)
-        if num_threads >= 1 and self.provider_type and len(text_batches) > 1:
-            with ThreadPoolExecutor(max_workers=num_threads) as executor:
-                future_to_batch = {
-                    executor.submit(process_batch, idx, batch): idx
-                    for idx, batch in enumerate(text_batches, start=1)
-                }
-
-                # Collect results in order
-                batch_results: list[tuple[int, list[Embedding]]] = []
-                for future in as_completed(future_to_batch):
-                    try:
-                        result = future.result()
-                        batch_results.append(result)
-                        if self.callback:
-                            self.callback.progress("_batch_encode_texts", 1)
-                    except Exception as e:
-                        logger.exception("Embedding model failed to process batch")
-                        raise e
-
-                # Sort by batch index and extend embeddings
-                batch_results.sort(key=lambda x: x[0])
-                for _, batch_embeddings in batch_results:
-                    embeddings.extend(batch_embeddings)
-        else:
-            # Original sequential processing
-            for idx, text_batch in enumerate(text_batches, start=1):
-                _, batch_embeddings = process_batch(idx, text_batch)
-                embeddings.extend(batch_embeddings)
-                if self.callback:
-                    self.callback.progress("_batch_encode_texts", 1)
+            embeddings.extend(response.embeddings)
 
+            if self.callback:
+                self.callback.progress("_batch_encode_texts", 1)
         return embeddings
 
     def encode(

backend/onyx/redis/redis_connector_doc_perm_sync.py

@@ -17,8 +17,6 @@ from onyx.redis.redis_pool import SCAN_ITER_COUNT_DEFAULT
 
 
 class RedisConnectorPermissionSyncPayload(BaseModel):
-    id: str
-    submitted: datetime
     started: datetime | None
     celery_task_id: str | None
 
@@ -43,12 +41,6 @@ class RedisConnectorPermissionSync:
     TASKSET_PREFIX = f"{PREFIX}_taskset"  # connectorpermissions_taskset
     SUBTASK_PREFIX = f"{PREFIX}+sub"  # connectorpermissions+sub
 
-    # used to signal the overall workflow is still active
-    # it's impossible to get the exact state of the system at a single point in time
-    # so we need a signal with a TTL to bridge gaps in our checks
-    ACTIVE_PREFIX = PREFIX + "_active"
-    ACTIVE_TTL = 3600
-
     def __init__(self, tenant_id: str | None, id: int, redis: redis.Redis) -> None:
         self.tenant_id: str | None = tenant_id
         self.id = id
@@ -62,7 +54,6 @@ class RedisConnectorPermissionSync:
         self.taskset_key = f"{self.TASKSET_PREFIX}_{id}"
 
         self.subtask_prefix: str = f"{self.SUBTASK_PREFIX}_{id}"
-        self.active_key = f"{self.ACTIVE_PREFIX}_{id}"
 
     def taskset_clear(self) -> None:
         self.redis.delete(self.taskset_key)
@@ -116,20 +107,6 @@ class RedisConnectorPermissionSync:
 
         self.redis.set(self.fence_key, payload.model_dump_json())
 
-    def set_active(self) -> None:
-        """This sets a signal to keep the permissioning flow from getting cleaned up within
-        the expiration time.
-
-        The slack in timing is needed to avoid race conditions where simply checking
-        the celery queue and task status could result in race conditions."""
-        self.redis.set(self.active_key, 0, ex=self.ACTIVE_TTL)
-
-    def active(self) -> bool:
-        if self.redis.exists(self.active_key):
-            return True
-
-        return False
-
     @property
     def generator_complete(self) -> int | None:
         """the fence payload is an int representing the starting number of
@@ -196,7 +173,6 @@ class RedisConnectorPermissionSync:
         return len(async_results)
 
     def reset(self) -> None:
-        self.redis.delete(self.active_key)
         self.redis.delete(self.generator_progress_key)
         self.redis.delete(self.generator_complete_key)
         self.redis.delete(self.taskset_key)
@@ -211,9 +187,6 @@ class RedisConnectorPermissionSync:
     @staticmethod
     def reset_all(r: redis.Redis) -> None:
         """Deletes all redis values for all connectors"""
-        for key in r.scan_iter(RedisConnectorPermissionSync.ACTIVE_PREFIX + "*"):
-            r.delete(key)
-
         for key in r.scan_iter(RedisConnectorPermissionSync.TASKSET_PREFIX + "*"):
             r.delete(key)
 

backend/onyx/redis/redis_connector_ext_group_sync.py

@@ -11,7 +11,6 @@ from onyx.redis.redis_pool import SCAN_ITER_COUNT_DEFAULT
 
 
 class RedisConnectorExternalGroupSyncPayload(BaseModel):
-    submitted: datetime
     started: datetime | None
     celery_task_id: str | None
 
@@ -136,12 +135,6 @@ class RedisConnectorExternalGroupSync:
     ) -> int | None:
         pass
 
-    def reset(self) -> None:
-        self.redis.delete(self.generator_progress_key)
-        self.redis.delete(self.generator_complete_key)
-        self.redis.delete(self.taskset_key)
-        self.redis.delete(self.fence_key)
-
     @staticmethod
     def remove_from_taskset(id: int, task_id: str, r: redis.Redis) -> None:
         taskset_key = f"{RedisConnectorExternalGroupSync.TASKSET_PREFIX}_{id}"

backend/onyx/redis/redis_connector_index.py

@@ -33,8 +33,8 @@ class RedisConnectorIndex:
     TERMINATE_TTL = 600
 
     # used to signal the overall workflow is still active
-    # it's impossible to get the exact state of the system at a single point in time
-    # so we need a signal with a TTL to bridge gaps in our checks
+    # there are gaps in time between states where we need some slack
+    # to correctly transition
     ACTIVE_PREFIX = PREFIX + "_active"
     ACTIVE_TTL = 3600
 

backend/onyx/redis/redis_pool.py

@@ -122,7 +122,7 @@ class TenantRedis(redis.Redis):
             "ttl",
         ]  # Regular methods that need simple prefixing
 
-        if item == "scan_iter" or item == "sscan_iter":
+        if item == "scan_iter":
             return self._prefix_scan_iter(original_attr)
         elif item in methods_to_wrap and callable(original_attr):
             return self._prefix_method(original_attr)

backend/onyx/server/documents/cc_pair.py

@@ -422,29 +422,27 @@ def sync_cc_pair(
     if redis_connector.permissions.fenced:
         raise HTTPException(
             status_code=HTTPStatus.CONFLICT,
-            detail="Permissions sync task already in progress.",
+            detail="Doc permissions sync task already in progress.",
         )
 
     logger.info(
-        f"Permissions sync cc_pair={cc_pair_id} "
+        f"Doc permissions sync cc_pair={cc_pair_id} "
         f"connector_id={cc_pair.connector_id} "
         f"credential_id={cc_pair.credential_id} "
         f"{cc_pair.connector.name} connector."
     )
-    payload_id = try_creating_permissions_sync_task(
+    tasks_created = try_creating_permissions_sync_task(
         primary_app, cc_pair_id, r, CURRENT_TENANT_ID_CONTEXTVAR.get()
     )
-    if not payload_id:
+    if not tasks_created:
         raise HTTPException(
             status_code=HTTPStatus.INTERNAL_SERVER_ERROR,
-            detail="Permissions sync task creation failed.",
+            detail="Doc permissions sync task creation failed.",
         )
 
-    logger.info(f"Permissions sync queued: cc_pair={cc_pair_id} id={payload_id}")
-
     return StatusResponse(
         success=True,
-        message="Successfully created the permissions sync task.",
+        message="Successfully created the doc permissions sync task.",
     )
 
 

backend/onyx/server/manage/models.py

@@ -44,11 +44,11 @@ class UserPreferences(BaseModel):
     chosen_assistants: list[int] | None = None
     hidden_assistants: list[int] = []
     visible_assistants: list[int] = []
+    recent_assistants: list[int] | None = None
     default_model: str | None = None
     auto_scroll: bool | None = None
     pinned_assistants: list[int] | None = None
     shortcut_enabled: bool | None = None
-    temperature_override_enabled: bool | None = None
 
 
 class UserInfo(BaseModel):
@@ -92,7 +92,6 @@ class UserInfo(BaseModel):
                     hidden_assistants=user.hidden_assistants,
                     pinned_assistants=user.pinned_assistants,
                     visible_assistants=user.visible_assistants,
-                    temperature_override_enabled=user.temperature_override_enabled,
                 )
             ),
             organization_name=organization_name,

backend/onyx/server/manage/users.py

@@ -568,9 +568,33 @@ def verify_user_logged_in(
 """APIs to adjust user preferences"""
 
 
-@router.patch("/temperature-override-enabled")
-def update_user_temperature_override_enabled(
-    temperature_override_enabled: bool,
+class ChosenDefaultModelRequest(BaseModel):
+    default_model: str | None = None
+
+
+class RecentAssistantsRequest(BaseModel):
+    current_assistant: int
+
+
+def update_recent_assistants(
+    recent_assistants: list[int] | None, current_assistant: int
+) -> list[int]:
+    if recent_assistants is None:
+        recent_assistants = []
+    else:
+        recent_assistants = [x for x in recent_assistants if x != current_assistant]
+
+    # Add current assistant to start of list
+    recent_assistants.insert(0, current_assistant)
+
+    # Keep only the 5 most recent assistants
+    recent_assistants = recent_assistants[:5]
+    return recent_assistants
+
+
+@router.patch("/user/recent-assistants")
+def update_user_recent_assistants(
+    request: RecentAssistantsRequest,
     user: User | None = Depends(current_user),
     db_session: Session = Depends(get_session),
 ) -> None:
@@ -578,26 +602,29 @@ def update_user_temperature_override_enabled(
         if AUTH_TYPE == AuthType.DISABLED:
             store = get_kv_store()
             no_auth_user = fetch_no_auth_user(store)
-            no_auth_user.preferences.temperature_override_enabled = (
-                temperature_override_enabled
+            preferences = no_auth_user.preferences
+            recent_assistants = preferences.recent_assistants
+            updated_preferences = update_recent_assistants(
+                recent_assistants, request.current_assistant
             )
-            set_no_auth_user_preferences(store, no_auth_user.preferences)
+            preferences.recent_assistants = updated_preferences
+            set_no_auth_user_preferences(store, preferences)
             return
         else:
             raise RuntimeError("This should never happen")
 
+    recent_assistants = UserInfo.from_model(user).preferences.recent_assistants
+    updated_recent_assistants = update_recent_assistants(
+        recent_assistants, request.current_assistant
+    )
     db_session.execute(
         update(User)
         .where(User.id == user.id)  # type: ignore
-        .values(temperature_override_enabled=temperature_override_enabled)
+        .values(recent_assistants=updated_recent_assistants)
     )
     db_session.commit()
 
 
-class ChosenDefaultModelRequest(BaseModel):
-    default_model: str | None = None
-
-
 @router.patch("/shortcut-enabled")
 def update_user_shortcut_enabled(
     shortcut_enabled: bool,
@@ -704,6 +731,30 @@ class ChosenAssistantsRequest(BaseModel):
     chosen_assistants: list[int]
 
 
+@router.patch("/user/assistant-list")
+def update_user_assistant_list(
+    request: ChosenAssistantsRequest,
+    user: User | None = Depends(current_user),
+    db_session: Session = Depends(get_session),
+) -> None:
+    if user is None:
+        if AUTH_TYPE == AuthType.DISABLED:
+            store = get_kv_store()
+            no_auth_user = fetch_no_auth_user(store)
+            no_auth_user.preferences.chosen_assistants = request.chosen_assistants
+            set_no_auth_user_preferences(store, no_auth_user.preferences)
+            return
+        else:
+            raise RuntimeError("This should never happen")
+
+    db_session.execute(
+        update(User)
+        .where(User.id == user.id)  # type: ignore
+        .values(chosen_assistants=request.chosen_assistants)
+    )
+    db_session.commit()
+
+
 def update_assistant_visibility(
     preferences: UserPreferences, assistant_id: int, show: bool
 ) -> UserPreferences:

backend/onyx/server/query_and_chat/chat_backend.py

@@ -18,6 +18,7 @@ from pydantic import BaseModel
 from sqlalchemy.orm import Session
 
 from onyx.auth.users import current_chat_accesssible_user
+from onyx.auth.users import current_limited_user
 from onyx.auth.users import current_user
 from onyx.chat.chat_utils import create_chat_chain
 from onyx.chat.chat_utils import extract_headers
@@ -77,7 +78,6 @@ from onyx.server.query_and_chat.models import LLMOverride
 from onyx.server.query_and_chat.models import PromptOverride
 from onyx.server.query_and_chat.models import RenameChatSessionResponse
 from onyx.server.query_and_chat.models import SearchFeedbackRequest
-from onyx.server.query_and_chat.models import UpdateChatSessionTemperatureRequest
 from onyx.server.query_and_chat.models import UpdateChatSessionThreadRequest
 from onyx.server.query_and_chat.token_limit import check_token_rate_limits
 from onyx.utils.headers import get_custom_tool_additional_request_headers
@@ -115,52 +115,12 @@ def get_user_chat_sessions(
                 shared_status=chat.shared_status,
                 folder_id=chat.folder_id,
                 current_alternate_model=chat.current_alternate_model,
-                current_temperature_override=chat.temperature_override,
             )
             for chat in chat_sessions
         ]
     )
 
 
-@router.put("/update-chat-session-temperature")
-def update_chat_session_temperature(
-    update_thread_req: UpdateChatSessionTemperatureRequest,
-    user: User | None = Depends(current_user),
-    db_session: Session = Depends(get_session),
-) -> None:
-    chat_session = get_chat_session_by_id(
-        chat_session_id=update_thread_req.chat_session_id,
-        user_id=user.id if user is not None else None,
-        db_session=db_session,
-    )
-
-    # Validate temperature_override
-    if update_thread_req.temperature_override is not None:
-        if (
-            update_thread_req.temperature_override < 0
-            or update_thread_req.temperature_override > 2
-        ):
-            raise HTTPException(
-                status_code=400, detail="Temperature must be between 0 and 2"
-            )
-
-        # Additional check for Anthropic models
-        if (
-            chat_session.current_alternate_model
-            and "anthropic" in chat_session.current_alternate_model.lower()
-        ):
-            if update_thread_req.temperature_override > 1:
-                raise HTTPException(
-                    status_code=400,
-                    detail="Temperature for Anthropic models must be between 0 and 1",
-                )
-
-    chat_session.temperature_override = update_thread_req.temperature_override
-
-    db_session.add(chat_session)
-    db_session.commit()
-
-
 @router.put("/update-chat-session-model")
 def update_chat_session_model(
     update_thread_req: UpdateChatSessionThreadRequest,
@@ -231,7 +191,6 @@ def get_chat_session(
         ],
         time_created=chat_session.time_created,
         shared_status=chat_session.shared_status,
-        current_temperature_override=chat_session.temperature_override,
     )
 
 
@@ -463,7 +422,7 @@ def set_message_as_latest(
 @router.post("/create-chat-message-feedback")
 def create_chat_feedback(
     feedback: ChatFeedbackRequest,
-    user: User | None = Depends(current_chat_accesssible_user),
+    user: User | None = Depends(current_limited_user),
     db_session: Session = Depends(get_session),
 ) -> None:
     user_id = user.id if user else None

backend/onyx/server/query_and_chat/models.py

@@ -42,11 +42,6 @@ class UpdateChatSessionThreadRequest(BaseModel):
     new_alternate_model: str
 
 
-class UpdateChatSessionTemperatureRequest(BaseModel):
-    chat_session_id: UUID
-    temperature_override: float
-
-
 class ChatSessionCreationRequest(BaseModel):
     # If not specified, use Onyx default persona
     persona_id: int = 0
@@ -113,10 +108,6 @@ class CreateChatMessageRequest(ChunkContext):
     llm_override: LLMOverride | None = None
     prompt_override: PromptOverride | None = None
 
-    # Allows the caller to override the temperature for the chat session
-    # this does persist in the chat thread details
-    temperature_override: float | None = None
-
     # allow user to specify an alternate assistnat
     alternate_assistant_id: int | None = None
 
@@ -177,7 +168,6 @@ class ChatSessionDetails(BaseModel):
     shared_status: ChatSessionSharedStatus
     folder_id: int | None = None
     current_alternate_model: str | None = None
-    current_temperature_override: float | None = None
 
 
 class ChatSessionsResponse(BaseModel):
@@ -241,7 +231,6 @@ class ChatSessionDetailResponse(BaseModel):
     time_created: datetime
     shared_status: ChatSessionSharedStatus
     current_alternate_model: str | None
-    current_temperature_override: float | None
 
 
 # This one is not used anymore

backend/onyx/server/utils.py

@@ -1,6 +1,4 @@
-import base64
 import json
-import os
 from datetime import datetime
 from typing import Any
 
@@ -68,10 +66,3 @@ def mask_credential_dict(credential_dict: dict[str, Any]) -> dict[str, str]:
         )
 
     return masked_creds
-
-
-def make_short_id() -> str:
-    """Fast way to generate a random 8 character id ... useful for tagging data
-    to trace it through a flow. This is definitely not guaranteed to be unique and is
-    targeted at the stated use case."""
-    return base64.b32encode(os.urandom(5)).decode("utf-8")[:8]  # 5 bytes → 8 chars

backend/onyx/setup.py

@@ -37,7 +37,6 @@ from onyx.document_index.vespa.index import VespaIndex
 from onyx.indexing.models import IndexingSetting
 from onyx.key_value_store.factory import get_kv_store
 from onyx.key_value_store.interface import KvKeyNotFoundError
-from onyx.llm.llm_provider_options import OPEN_AI_MODEL_NAMES
 from onyx.natural_language_processing.search_nlp_models import EmbeddingModel
 from onyx.natural_language_processing.search_nlp_models import warm_up_bi_encoder
 from onyx.natural_language_processing.search_nlp_models import warm_up_cross_encoder
@@ -280,7 +279,6 @@ def setup_postgres(db_session: Session) -> None:
     if GEN_AI_API_KEY and fetch_default_provider(db_session) is None:
         # Only for dev flows
         logger.notice("Setting up default OpenAI LLM for dev.")
-
         llm_model = GEN_AI_MODEL_VERSION or "gpt-4o-mini"
         fast_model = FAST_GEN_AI_MODEL_VERSION or "gpt-4o-mini"
         model_req = LLMProviderUpsertRequest(
@@ -294,8 +292,8 @@ def setup_postgres(db_session: Session) -> None:
             fast_default_model_name=fast_model,
             is_public=True,
             groups=[],
-            display_model_names=OPEN_AI_MODEL_NAMES,
-            model_names=OPEN_AI_MODEL_NAMES,
+            display_model_names=[llm_model, fast_model],
+            model_names=[llm_model, fast_model],
         )
         new_llm_provider = upsert_llm_provider(
             llm_provider=model_req, db_session=db_session

backend/onyx/utils/logger.py

@@ -26,13 +26,6 @@ doc_permission_sync_ctx: contextvars.ContextVar[
 ] = contextvars.ContextVar("doc_permission_sync_ctx", default=dict())
 
 
-class LoggerContextVars:
-    @staticmethod
-    def reset() -> None:
-        pruning_ctx.set(dict())
-        doc_permission_sync_ctx.set(dict())
-
-
 class TaskAttemptSingleton:
     """Used to tell if this process is an indexing job, and if so what is the
     unique identifier for this indexing attempt. For things like the API server,
@@ -77,32 +70,27 @@ class OnyxLoggingAdapter(logging.LoggerAdapter):
     ) -> tuple[str, MutableMapping[str, Any]]:
         # If this is an indexing job, add the attempt ID to the log message
         # This helps filter the logs for this specific indexing
-        while True:
-            pruning_ctx_dict = pruning_ctx.get()
-            if len(pruning_ctx_dict) > 0:
-                if "request_id" in pruning_ctx_dict:
-                    msg = f"[Prune: {pruning_ctx_dict['request_id']}] {msg}"
+        index_attempt_id = TaskAttemptSingleton.get_index_attempt_id()
+        cc_pair_id = TaskAttemptSingleton.get_connector_credential_pair_id()
 
-                if "cc_pair_id" in pruning_ctx_dict:
-                    msg = f"[CC Pair: {pruning_ctx_dict['cc_pair_id']}] {msg}"
-                break
-
-            doc_permission_sync_ctx_dict = doc_permission_sync_ctx.get()
-            if len(doc_permission_sync_ctx_dict) > 0:
-                if "request_id" in doc_permission_sync_ctx_dict:
-                    msg = f"[Doc Permissions Sync: {doc_permission_sync_ctx_dict['request_id']}] {msg}"
-                break
-
-            index_attempt_id = TaskAttemptSingleton.get_index_attempt_id()
-            cc_pair_id = TaskAttemptSingleton.get_connector_credential_pair_id()
+        doc_permission_sync_ctx_dict = doc_permission_sync_ctx.get()
+        pruning_ctx_dict = pruning_ctx.get()
+        if len(pruning_ctx_dict) > 0:
+            if "request_id" in pruning_ctx_dict:
+                msg = f"[Prune: {pruning_ctx_dict['request_id']}] {msg}"
 
+            if "cc_pair_id" in pruning_ctx_dict:
+                msg = f"[CC Pair: {pruning_ctx_dict['cc_pair_id']}] {msg}"
+        elif len(doc_permission_sync_ctx_dict) > 0:
+            if "request_id" in doc_permission_sync_ctx_dict:
+                msg = f"[Doc Permissions Sync: {doc_permission_sync_ctx_dict['request_id']}] {msg}"
+        else:
             if index_attempt_id is not None:
                 msg = f"[Index Attempt: {index_attempt_id}] {msg}"
 
             if cc_pair_id is not None:
                 msg = f"[CC Pair: {cc_pair_id}] {msg}"
 
-            break
         # Add tenant information if it differs from default
         # This will always be the case for authenticated API requests
         if MULTI_TENANT:

backend/onyx/utils/telemetry.py

@@ -1,4 +1,3 @@
-import contextvars
 import threading
 import uuid
 from enum import Enum
@@ -42,7 +41,7 @@ def _get_or_generate_customer_id_mt(tenant_id: str) -> str:
     return str(uuid.uuid5(uuid.NAMESPACE_X500, tenant_id))
 
 
-def get_or_generate_uuid() -> str:
+def get_or_generate_uuid(tenant_id: str | None) -> str:
     # TODO: split out the whole "instance UUID" generation logic into a separate
     # utility function. Telemetry should not be aware at all of how the UUID is
     # generated/stored.
@@ -53,7 +52,7 @@ def get_or_generate_uuid() -> str:
     if _CACHED_UUID is not None:
         return _CACHED_UUID
 
-    kv_store = get_kv_store()
+    kv_store = get_kv_store(tenant_id=tenant_id)
 
     try:
         _CACHED_UUID = cast(str, kv_store.load(KV_CUSTOMER_UUID_KEY))
@@ -64,18 +63,18 @@ def get_or_generate_uuid() -> str:
     return _CACHED_UUID
 
 
-def _get_or_generate_instance_domain() -> str | None:  #
+def _get_or_generate_instance_domain(tenant_id: str | None = None) -> str | None:  #
     global _CACHED_INSTANCE_DOMAIN
 
     if _CACHED_INSTANCE_DOMAIN is not None:
         return _CACHED_INSTANCE_DOMAIN
 
-    kv_store = get_kv_store()
+    kv_store = get_kv_store(tenant_id=tenant_id)
 
     try:
         _CACHED_INSTANCE_DOMAIN = cast(str, kv_store.load(KV_INSTANCE_DOMAIN_KEY))
     except KvKeyNotFoundError:
-        with get_session_with_tenant() as db_session:
+        with get_session_with_tenant(tenant_id=tenant_id) as db_session:
             first_user = db_session.query(User).first()
             if first_user:
                 _CACHED_INSTANCE_DOMAIN = first_user.email.split("@")[-1]
@@ -104,7 +103,7 @@ def optional_telemetry(
                 customer_uuid = (
                     _get_or_generate_customer_id_mt(tenant_id)
                     if MULTI_TENANT
-                    else get_or_generate_uuid()
+                    else get_or_generate_uuid(tenant_id)
                 )
                 payload = {
                     "data": data,
@@ -116,7 +115,9 @@ def optional_telemetry(
                     "is_cloud": MULTI_TENANT,
                 }
                 if ENTERPRISE_EDITION_ENABLED:
-                    payload["instance_domain"] = _get_or_generate_instance_domain()
+                    payload["instance_domain"] = _get_or_generate_instance_domain(
+                        tenant_id
+                    )
                 requests.post(
                     _DANSWER_TELEMETRY_ENDPOINT,
                     headers={"Content-Type": "application/json"},
@@ -127,12 +128,8 @@ def optional_telemetry(
                 # This way it silences all thread level logging as well
                 pass
 
-        # Run in separate thread with the same context as the current thread
-        # This is to ensure that the thread gets the current tenant ID
-        current_context = contextvars.copy_context()
-        thread = threading.Thread(
-            target=lambda: current_context.run(telemetry_logic), daemon=True
-        )
+        # Run in separate thread to have minimal overhead in main flows
+        thread = threading.Thread(target=telemetry_logic, daemon=True)
         thread.start()
     except Exception:
         # Should never interfere with normal functions of Onyx

backend/requirements/default.txt

@@ -81,7 +81,6 @@ hubspot-api-client==8.1.0
 asana==5.0.8
 dropbox==11.36.2
 boto3-stubs[s3]==1.34.133
-shapely==2.0.6
 stripe==10.12.0
 urllib3==2.2.3
 mistune==0.8.4

backend/shared_configs/contextvars.py

@@ -6,13 +6,3 @@ from shared_configs.configs import POSTGRES_DEFAULT_SCHEMA
 CURRENT_TENANT_ID_CONTEXTVAR = contextvars.ContextVar(
     "current_tenant_id", default=POSTGRES_DEFAULT_SCHEMA
 )
-
-
-"""Utils related to contextvars"""
-
-
-def get_current_tenant_id() -> str:
-    tenant_id = CURRENT_TENANT_ID_CONTEXTVAR.get()
-    if tenant_id is None:
-        raise RuntimeError("Tenant ID is not set. This should never happen.")
-    return tenant_id

backend/tests/unit/onyx/llm/test_chat_llm.py

@@ -9,7 +9,6 @@ from litellm.types.utils import ChatCompletionDeltaToolCall
 from litellm.types.utils import Delta
 from litellm.types.utils import Function as LiteLLMFunction
 
-from onyx.configs.app_configs import MOCK_LLM_RESPONSE
 from onyx.llm.chat_llm import DefaultMultiLLM
 
 
@@ -144,7 +143,6 @@ def test_multiple_tool_calls(default_multi_llm: DefaultMultiLLM) -> None:
             temperature=0.0,  # Default value from GEN_AI_TEMPERATURE
             timeout=30,
             parallel_tool_calls=False,
-            mock_response=MOCK_LLM_RESPONSE,
         )
 
 
@@ -289,5 +287,4 @@ def test_multiple_tool_calls_streaming(default_multi_llm: DefaultMultiLLM) -> No
             temperature=0.0,  # Default value from GEN_AI_TEMPERATURE
             timeout=30,
             parallel_tool_calls=False,
-            mock_response=MOCK_LLM_RESPONSE,
         )

deployment/cloud_kubernetes/hpa/workers_hpa.yaml

@@ -0,0 +1,75 @@
+apiVersion: autoscaling/v2
+kind: HorizontalPodAutoscaler
+metadata:
+  name: celery-worker-heavy-hpa
+spec:
+  scaleTargetRef:
+    apiVersion: apps/v1
+    kind: Deployment
+    name: celery-worker-heavy
+  minReplicas: 1
+  maxReplicas: 5
+  metrics:
+    - type: Resource
+      resource:
+        name: cpu
+        target:
+          type: Utilization
+          averageUtilization: 60
+---
+apiVersion: autoscaling/v2
+kind: HorizontalPodAutoscaler
+metadata:
+  name: celery-worker-light-hpa
+spec:
+  scaleTargetRef:
+    apiVersion: apps/v1
+    kind: Deployment
+    name: celery-worker-light
+  minReplicas: 1
+  maxReplicas: 10
+  metrics:
+    - type: Resource
+      resource:
+        name: cpu
+        target:
+          type: Utilization
+          averageUtilization: 70
+---
+apiVersion: autoscaling/v2
+kind: HorizontalPodAutoscaler
+metadata:
+  name: celery-worker-indexing-hpa
+spec:
+  scaleTargetRef:
+    apiVersion: apps/v1
+    kind: Deployment
+    name: celery-worker-indexing
+  minReplicas: 1
+  maxReplicas: 10
+  metrics:
+    - type: Resource
+      resource:
+        name: cpu
+        target:
+          type: Utilization
+          averageUtilization: 70
+---
+apiVersion: autoscaling/v2
+kind: HorizontalPodAutoscaler
+metadata:
+  name: celery-worker-monitoring-hpa
+spec:
+  scaleTargetRef:
+    apiVersion: apps/v1
+    kind: Deployment
+    name: celery-worker-indexing
+  minReplicas: 1
+  maxReplicas: 4
+  metrics:
+    - type: Resource
+      resource:
+        name: cpu
+        target:
+          type: Utilization
+          averageUtilization: 70

deployment/cloud_kubernetes/keda/celery-worker-auth.yaml

@@ -0,0 +1,13 @@
+apiVersion: keda.sh/v1alpha1
+kind: TriggerAuthentication
+metadata:
+  name: celery-worker-auth
+  namespace: onyx
+spec:
+  secretTargetRef:
+    - parameter: host
+      name: keda-redis-secret
+      key: host
+    - parameter: password
+      name: keda-redis-secret
+      key: password

deployment/cloud_kubernetes/keda/celery-worker-indexing-scaledobject.yaml

@@ -0,0 +1,53 @@
+apiVersion: keda.sh/v1alpha1
+kind: ScaledObject
+metadata:
+  name: celery-worker-indexing-scaledobject
+  namespace: onyx
+  labels:
+    app: celery-worker-indexing
+spec:
+  scaleTargetRef:
+    name: celery-worker-indexing
+  minReplicaCount: 1
+  maxReplicaCount: 30
+  triggers:
+    - type: redis
+      metadata:
+        sslEnabled: "true"
+        port: "6379"
+        enableTLS: "true"
+        listName: connector_indexing
+        listLength: "1"
+        databaseIndex: "15"
+      authenticationRef:
+        name: celery-worker-auth
+
+    - type: redis
+      metadata:
+        sslEnabled: "true"
+        port: "6379"
+        enableTLS: "true"
+        listName: connector_indexing:2
+        listLength: "1"
+        databaseIndex: "15"
+      authenticationRef:
+        name: celery-worker-auth
+    - type: redis
+      metadata:
+        sslEnabled: "true"
+        port: "6379"
+        enableTLS: "true"
+        listName: connector_indexing:3
+        listLength: "1"
+        databaseIndex: "15"
+      authenticationRef:
+        name: celery-worker-auth
+    - type: cpu
+      metadata:
+        type: Utilization
+        value: "70"
+
+    - type: memory
+      metadata:
+        type: Utilization
+        value: "70"

deployment/cloud_kubernetes/keda/celery-worker-light-scaledobject.yaml

@@ -0,0 +1,58 @@
+apiVersion: keda.sh/v1alpha1
+kind: ScaledObject
+metadata:
+  name: celery-worker-light-scaledobject
+  namespace: onyx
+  labels:
+    app: celery-worker-light
+spec:
+  scaleTargetRef:
+    name: celery-worker-light
+  minReplicaCount: 5
+  maxReplicaCount: 20
+  triggers:
+    - type: redis
+      metadata:
+        port: "6379"
+        enableTLS: "true"
+        listName: vespa_metadata_sync
+        listLength: "1"
+        databaseIndex: "15"
+      authenticationRef:
+        name: celery-worker-auth
+    - type: redis
+      metadata:
+        port: "6379"
+        enableTLS: "true"
+        listName: vespa_metadata_sync:2
+        listLength: "1"
+        databaseIndex: "15"
+      authenticationRef:
+        name: celery-worker-auth
+    - type: redis
+      metadata:
+        port: "6379"
+        enableTLS: "true"
+        listName: vespa_metadata_sync:3
+        listLength: "1"
+        databaseIndex: "15"
+      authenticationRef:
+        name: celery-worker-auth
+    - type: redis
+      metadata:
+        port: "6379"
+        enableTLS: "true"
+        listName: connector_deletion
+        listLength: "1"
+        databaseIndex: "15"
+      authenticationRef:
+        name: celery-worker-auth
+    - type: redis
+      metadata:
+        port: "6379"
+        enableTLS: "true"
+        listName: connector_deletion:2
+        listLength: "1"
+        databaseIndex: "15"
+      authenticationRef:
+        name: celery-worker-auth

deployment/cloud_kubernetes/keda/celery-worker-primary-scaledobject.yaml

@@ -0,0 +1,70 @@
+apiVersion: keda.sh/v1alpha1
+kind: ScaledObject
+metadata:
+  name: celery-worker-primary-scaledobject
+  namespace: onyx
+  labels:
+    app: celery-worker-primary
+spec:
+  scaleTargetRef:
+    name: celery-worker-primary
+  pollingInterval: 15 # Check every 15 seconds
+  cooldownPeriod: 30 # Wait 30 seconds before scaling down
+  minReplicaCount: 4
+  maxReplicaCount: 4
+  triggers:
+    - type: redis
+      metadata:
+        port: "6379"
+        enableTLS: "true"
+        listName: celery
+        listLength: "1"
+        databaseIndex: "15"
+      authenticationRef:
+        name: celery-worker-auth
+
+    - type: redis
+      metadata:
+        port: "6379"
+        enableTLS: "true"
+        listName: celery:1
+        listLength: "1"
+        databaseIndex: "15"
+      authenticationRef:
+        name: celery-worker-auth
+    - type: redis
+      metadata:
+        port: "6379"
+        enableTLS: "true"
+        listName: celery:2
+        listLength: "1"
+        databaseIndex: "15"
+      authenticationRef:
+        name: celery-worker-auth
+    - type: redis
+      metadata:
+        port: "6379"
+        enableTLS: "true"
+        listName: celery:3
+        listLength: "1"
+        databaseIndex: "15"
+      authenticationRef:
+        name: celery-worker-auth
+    - type: redis
+      metadata:
+        port: "6379"
+        enableTLS: "true"
+        listName: periodic_tasks
+        listLength: "1"
+        databaseIndex: "15"
+      authenticationRef:
+        name: celery-worker-auth
+    - type: redis
+      metadata:
+        port: "6379"
+        enableTLS: "true"
+        listName: periodic_tasks:2
+        listLength: "1"
+        databaseIndex: "15"
+      authenticationRef:
+        name: celery-worker-auth

deployment/cloud_kubernetes/keda/indexing-model-server-scaledobject.yaml

@@ -0,0 +1,19 @@
+apiVersion: keda.sh/v1alpha1
+kind: ScaledObject
+metadata:
+  name: indexing-model-server-scaledobject
+  namespace: onyx
+  labels:
+    app: indexing-model-server
+spec:
+  scaleTargetRef:
+    name: indexing-model-server-deployment
+  pollingInterval: 15 # Check every 15 seconds
+  cooldownPeriod: 30 # Wait 30 seconds before scaling down
+  minReplicaCount: 10
+  maxReplicaCount: 10
+  triggers:
+    - type: cpu
+      metadata:
+        type: Utilization
+        value: "70"

deployment/cloud_kubernetes/keda/keda-redis-secret.yaml

@@ -0,0 +1,9 @@
+apiVersion: v1
+kind: Secret
+metadata:
+  name: keda-redis-secret
+  namespace: onyx
+type: Opaque
+data:
+  host: { base64 encoded host here }
+  password: { base64 encoded password here }

deployment/cloud_kubernetes/workers/beat.yaml

@@ -0,0 +1,44 @@
+apiVersion: apps/v1
+kind: Deployment
+metadata:
+  name: celery-beat
+spec:
+  replicas: 1
+  selector:
+    matchLabels:
+      app: celery-beat
+  template:
+    metadata:
+      labels:
+        app: celery-beat
+    spec:
+      containers:
+        - name: celery-beat
+          image: onyxdotapp/onyx-backend-cloud:v0.14.0-cloud.beta.21
+          imagePullPolicy: IfNotPresent
+          command:
+            [
+              "celery",
+              "-A",
+              "onyx.background.celery.versioned_apps.beat",
+              "beat",
+              "--loglevel=INFO",
+            ]
+          env:
+            - name: REDIS_PASSWORD
+              valueFrom:
+                secretKeyRef:
+                  name: onyx-secrets
+                  key: redis_password
+            - name: ONYX_VERSION
+              value: "v0.11.0-cloud.beta.8"
+          envFrom:
+            - configMapRef:
+                name: env-configmap
+          resources:
+            requests:
+              cpu: "250m"
+              memory: "512Mi"
+            limits:
+              cpu: "500m"
+              memory: "1Gi"

deployment/cloud_kubernetes/workers/heavy_worker.yaml

@@ -0,0 +1,60 @@
+apiVersion: apps/v1
+kind: Deployment
+metadata:
+  name: celery-worker-heavy
+spec:
+  replicas: 2
+  selector:
+    matchLabels:
+      app: celery-worker-heavy
+  template:
+    metadata:
+      labels:
+        app: celery-worker-heavy
+    spec:
+      containers:
+        - name: celery-worker-heavy
+          image: onyxdotapp/onyx-backend-cloud:v0.14.0-cloud.beta.21
+          imagePullPolicy: IfNotPresent
+          command:
+            [
+              "celery",
+              "-A",
+              "onyx.background.celery.versioned_apps.heavy",
+              "worker",
+              "--loglevel=INFO",
+              "--hostname=heavy@%n",
+              "-Q",
+              "connector_pruning,connector_doc_permissions_sync,connector_external_group_sync",
+            ]
+          env:
+            - name: REDIS_PASSWORD
+              valueFrom:
+                secretKeyRef:
+                  name: onyx-secrets
+                  key: redis_password
+            - name: ONYX_VERSION
+              value: "v0.11.0-cloud.beta.8"
+          envFrom:
+            - configMapRef:
+                name: env-configmap
+          volumeMounts:
+            - name: vespa-certificates
+              mountPath: "/app/certs"
+              readOnly: true
+          resources:
+            requests:
+              cpu: "1000m"
+              memory: "2Gi"
+            limits:
+              cpu: "2000m"
+              memory: "4Gi"
+      volumes:
+        - name: vespa-certificates
+          secret:
+            secretName: vespa-certificates
+            items:
+              - key: cert.pem
+                path: cert.pem
+              - key: key.pem
+                path: key.pem

deployment/cloud_kubernetes/workers/indexing_worker.yaml

@@ -0,0 +1,62 @@
+apiVersion: apps/v1
+kind: Deployment
+metadata:
+  name: celery-worker-indexing
+spec:
+  replicas: 1
+  selector:
+    matchLabels:
+      app: celery-worker-indexing
+  template:
+    metadata:
+      labels:
+        app: celery-worker-indexing
+    spec:
+      containers:
+        - name: celery-worker-indexing
+          image: onyxdotapp/onyx-backend-cloud:v0.14.0-cloud.beta.21
+          imagePullPolicy: IfNotPresent
+          command:
+            [
+              "celery",
+              "-A",
+              "onyx.background.celery.versioned_apps.indexing",
+              "worker",
+              "--loglevel=INFO",
+              "--hostname=indexing@%n",
+              "-Q",
+              "connector_indexing",
+              "--prefetch-multiplier=1",
+              "--concurrency=10",
+            ]
+          env:
+            - name: REDIS_PASSWORD
+              valueFrom:
+                secretKeyRef:
+                  name: onyx-secrets
+                  key: redis_password
+            - name: ONYX_VERSION
+              value: "v0.11.0-cloud.beta.8"
+          envFrom:
+            - configMapRef:
+                name: env-configmap
+          volumeMounts:
+            - name: vespa-certificates
+              mountPath: "/app/certs"
+              readOnly: true
+          resources:
+            requests:
+              cpu: "500m"
+              memory: "4Gi"
+            limits:
+              cpu: "1000m"
+              memory: "8Gi"
+      volumes:
+        - name: vespa-certificates
+          secret:
+            secretName: vespa-certificates
+            items:
+              - key: cert.pem
+                path: cert.pem
+              - key: key.pem
+                path: key.pem

deployment/cloud_kubernetes/workers/light_worker.yaml

@@ -0,0 +1,62 @@
+apiVersion: apps/v1
+kind: Deployment
+metadata:
+  name: celery-worker-light
+spec:
+  replicas: 1
+  selector:
+    matchLabels:
+      app: celery-worker-light
+  template:
+    metadata:
+      labels:
+        app: celery-worker-light
+    spec:
+      containers:
+        - name: celery-worker-light
+          image: onyxdotapp/onyx-backend-cloud:v0.14.0-cloud.beta.21
+          imagePullPolicy: IfNotPresent
+          command:
+            [
+              "celery",
+              "-A",
+              "onyx.background.celery.versioned_apps.light",
+              "worker",
+              "--loglevel=INFO",
+              "--hostname=light@%n",
+              "-Q",
+              "vespa_metadata_sync,connector_deletion,doc_permissions_upsert",
+              "--prefetch-multiplier=1",
+              "--concurrency=10",
+            ]
+          env:
+            - name: REDIS_PASSWORD
+              valueFrom:
+                secretKeyRef:
+                  name: onyx-secrets
+                  key: redis_password
+            - name: ONYX_VERSION
+              value: "v0.11.0-cloud.beta.8"
+          envFrom:
+            - configMapRef:
+                name: env-configmap
+          volumeMounts:
+            - name: vespa-certificates
+              mountPath: "/app/certs"
+              readOnly: true
+          resources:
+            requests:
+              cpu: "500m"
+              memory: "1Gi"
+            limits:
+              cpu: "1000m"
+              memory: "2Gi"
+      volumes:
+        - name: vespa-certificates
+          secret:
+            secretName: vespa-certificates
+            items:
+              - key: cert.pem
+                path: cert.pem
+              - key: key.pem
+                path: key.pem

deployment/cloud_kubernetes/workers/monitoring.yaml

@@ -0,0 +1,62 @@
+apiVersion: apps/v1
+kind: Deployment
+metadata:
+  name: celery-worker-monitoring
+spec:
+  replicas: 2
+  selector:
+    matchLabels:
+      app: celery-worker-monitoring
+  template:
+    metadata:
+      labels:
+        app: celery-worker-monitoring
+    spec:
+      containers:
+        - name: celery-worker-monitoring
+          image: onyxdotapp/onyx-backend-cloud:v0.14.0-cloud.beta.21
+          imagePullPolicy: IfNotPresent
+          command:
+            [
+              "celery",
+              "-A",
+              "onyx.background.celery.versioned_apps.monitoring",
+              "worker",
+              "--loglevel=INFO",
+              "--hostname=monitoring@%n",
+              "-Q",
+              "monitoring",
+              "--prefetch-multiplier=8",
+              "--concurrency=8",
+            ]
+          env:
+            - name: REDIS_PASSWORD
+              valueFrom:
+                secretKeyRef:
+                  name: onyx-secrets
+                  key: redis_password
+            - name: ONYX_VERSION
+              value: "v0.11.0-cloud.beta.8"
+          envFrom:
+            - configMapRef:
+                name: env-configmap
+          volumeMounts:
+            - name: vespa-certificates
+              mountPath: "/app/certs"
+              readOnly: true
+          resources:
+            requests:
+              cpu: "1000m"
+              memory: "1Gi"
+            limits:
+              cpu: "1000m"
+              memory: "1Gi"
+      volumes:
+        - name: vespa-certificates
+          secret:
+            secretName: vespa-certificates
+            items:
+              - key: cert.pem
+                path: cert.pem
+              - key: key.pem
+                path: key.pem

deployment/cloud_kubernetes/workers/primary.yaml

@@ -0,0 +1,62 @@
+apiVersion: apps/v1
+kind: Deployment
+metadata:
+  name: celery-worker-primary
+spec:
+  replicas: 1
+  selector:
+    matchLabels:
+      app: celery-worker-primary
+  template:
+    metadata:
+      labels:
+        app: celery-worker-primary
+    spec:
+      containers:
+        - name: celery-worker-primary
+          image: onyxdotapp/onyx-backend-cloud:v0.14.0-cloud.beta.21
+          imagePullPolicy: IfNotPresent
+          command:
+            [
+              "celery",
+              "-A",
+              "onyx.background.celery.versioned_apps.primary",
+              "worker",
+              "--loglevel=INFO",
+              "--hostname=primary@%n",
+              "-Q",
+              "celery,periodic_tasks",
+              "--prefetch-multiplier=1",
+              "--concurrency=10",
+            ]
+          env:
+            - name: REDIS_PASSWORD
+              valueFrom:
+                secretKeyRef:
+                  name: onyx-secrets
+                  key: redis_password
+            - name: ONYX_VERSION
+              value: "v0.11.0-cloud.beta.8"
+          envFrom:
+            - configMapRef:
+                name: env-configmap
+          volumeMounts:
+            - name: vespa-certificates
+              mountPath: "/app/certs"
+              readOnly: true
+          resources:
+            requests:
+              cpu: "500m"
+              memory: "1Gi"
+            limits:
+              cpu: "1000m"
+              memory: "2Gi"
+      volumes:
+        - name: vespa-certificates
+          secret:
+            secretName: vespa-certificates
+            items:
+              - key: cert.pem
+                path: cert.pem
+              - key: key.pem
+                path: key.pem

deployment/helm/charts/onyx/Chart.yaml

@@ -6,7 +6,7 @@ sources:
   - "https://github.com/onyx-dot-app/onyx"
 type: application
 version: 0.2.1
-appVersion: latest
+appVersion: "latest"
 annotations:
   category: Productivity
   licenses: MIT

deployment/helm/charts/onyx/templates/api-deployment.yaml

@@ -45,10 +45,10 @@ spec:
             - |
               alembic upgrade head &&
               echo "Starting Onyx Api Server" &&
-              uvicorn onyx.main:app --host 0.0.0.0 --port {{ .Values.api.containerPorts.server }}
+              uvicorn onyx.main:app --host 0.0.0.0 --port 8080
           ports:
             - name: api-server-port
-              containerPort: {{ .Values.api.containerPorts.server }}
+              containerPort: {{ .Values.api.service.port }}
               protocol: TCP
           resources:
             {{- toYaml .Values.api.resources | nindent 12 }}

deployment/helm/charts/onyx/templates/api-service.yaml

@@ -11,10 +11,10 @@ metadata:
 spec:
   type: {{ .Values.api.service.type }}
   ports:
-    - port: {{ .Values.api.service.servicePort }}
-      targetPort: {{ .Values.api.service.targetPort }}
+    - port: {{ .Values.api.service.port }}
+      targetPort: api-server-port
       protocol: TCP
-      name: {{ .Values.api.service.portName }}
+      name: api-server-port
   selector:
     {{- include "onyx-stack.selectorLabels" . | nindent 4 }}
     {{- if .Values.api.deploymentLabels }}

deployment/helm/charts/onyx/templates/indexing-model-deployment.yaml

@@ -5,7 +5,7 @@ metadata:
   labels:
     {{- include "onyx-stack.labels" . | nindent 4 }}
 spec:
-  replicas: {{ .Values.indexCapability.replicaCount }}
+  replicas: 1
   selector:
     matchLabels:
       {{- include "onyx-stack.selectorLabels" . | nindent 6 }}
@@ -25,14 +25,12 @@ spec:
         {{- end }}
     spec:
       containers:
-      - name: {{ .Values.indexCapability.name }}
-        image: "{{ .Values.indexCapability.image.repository }}:{{ .Values.indexCapability.image.tag | default .Chart.AppVersion }}"           
-        imagePullPolicy: {{ .Values.indexCapability.image.pullPolicy }}
-        command: [ "uvicorn", "model_server.main:app", "--host", "0.0.0.0", "--port", "{{ .Values.indexCapability.containerPorts.server }}", "--limit-concurrency", "{{ .Values.indexCapability.limitConcurrency }}" ]
+      - name: indexing-model-server
+        image: onyxdotapp/onyx-model-server:latest
+        imagePullPolicy: IfNotPresent
+        command: [ "uvicorn", "model_server.main:app", "--host", "0.0.0.0", "--port", "9000", "--limit-concurrency", "10" ]
         ports:
-        - name: model-server
-          containerPort: {{ .Values.indexCapability.containerPorts.server }}
-          protocol: TCP
+        - containerPort: 9000
         envFrom:
           - configMapRef:
               name: {{ .Values.config.envConfigMapName }}

deployment/helm/charts/onyx/templates/indexing-model-pvc.yaml

@@ -3,9 +3,8 @@ kind: PersistentVolumeClaim
 metadata:
   name: {{ .Values.indexCapability.indexingModelPVC.name }}
 spec:
-  storageClassName: {{ .Values.persistent.storageClassName }}
   accessModes:
     - {{ .Values.indexCapability.indexingModelPVC.accessMode | quote }}
   resources:
     requests:
-      storage: {{ .Values.indexCapability.indexingModelPVC.storage | quote }}
+      storage: {{ .Values.indexCapability.indexingModelPVC.storage | quote }}

deployment/helm/charts/onyx/templates/indexing-model-service.yaml

@@ -11,8 +11,8 @@ spec:
     {{- toYaml .Values.indexCapability.deploymentLabels | nindent 4 }}
     {{- end }}
   ports:
-    - name: {{ .Values.indexCapability.service.portName }}
+    - name: {{ .Values.indexCapability.service.name }}
       protocol: TCP
-      port: {{ .Values.indexCapability.service.servicePort  }}
-      targetPort: {{ .Values.indexCapability.service.targetPort }}
-  type: {{ .Values.indexCapability.service.type }}
+      port: {{ .Values.indexCapability.service.port }}
+      targetPort: {{ .Values.indexCapability.service.port }}
+  type: {{ .Values.indexCapability.service.type }}

deployment/helm/charts/onyx/templates/inference-model-deployment.yaml

@@ -3,14 +3,14 @@ kind: Deployment
 metadata:
   name: {{ include "onyx-stack.fullname" . }}-inference-model
   labels:
-    {{- range .Values.inferenceCapability.labels }}
+    {{- range .Values.inferenceCapability.deployment.labels }}
     {{ .key }}: {{ .value }}
     {{- end }}
 spec:
-  replicas: {{ .Values.inferenceCapability.replicaCount }}
+  replicas: {{ .Values.inferenceCapability.deployment.replicas }}
   selector:
     matchLabels:
-      {{- range .Values.inferenceCapability.labels }}
+      {{- range .Values.inferenceCapability.deployment.labels }}
       {{ .key }}: {{ .value }}
       {{- end }}
   template:
@@ -21,26 +21,24 @@ spec:
         {{- end }}
     spec:
       containers:
-      - name: model-server-inference
-        image: "{{ .Values.inferenceCapability.image.repository }}:{{ .Values.inferenceCapability.image.tag | default .Chart.AppVersion }}"
-        imagePullPolicy: {{ .Values.inferenceCapability.image.pullPolicy }}
-        command: [ "uvicorn", "model_server.main:app", "--host", "0.0.0.0", "--port", "{{ .Values.inferenceCapability.containerPorts.server }}" ]
+      - name: {{ .Values.inferenceCapability.service.name }}
+        image: {{ .Values.inferenceCapability.deployment.image.repository }}:{{ .Values.inferenceCapability.deployment.image.tag }}
+        imagePullPolicy: {{ .Values.inferenceCapability.deployment.image.pullPolicy }}
+        command: {{ toYaml .Values.inferenceCapability.deployment.command | nindent 14 }}
         ports:
-        - name: model-server
-          containerPort: {{ .Values.inferenceCapability.containerPorts.server }}
-          protocol: TCP
+        - containerPort: {{ .Values.inferenceCapability.service.port }}
         envFrom:
         - configMapRef:
             name: {{ .Values.config.envConfigMapName }}
         env:
           {{- include "onyx-stack.envSecrets" . | nindent 12}}
         volumeMounts:
-        {{- range .Values.inferenceCapability.volumeMounts }}
+        {{- range .Values.inferenceCapability.deployment.volumeMounts }}
         - name: {{ .name }}
           mountPath: {{ .mountPath }}
         {{- end }}
       volumes:
-      {{- range .Values.inferenceCapability.volumes }}
+      {{- range .Values.inferenceCapability.deployment.volumes }}
       - name: {{ .name }}
         persistentVolumeClaim:
           claimName: {{ .persistentVolumeClaim.claimName }}

deployment/helm/charts/onyx/templates/inference-model-pvc.yaml

@@ -3,7 +3,6 @@ kind: PersistentVolumeClaim
 metadata:
   name: {{ .Values.inferenceCapability.pvc.name }}
 spec:
-  storageClassName: {{ .Values.persistent.storageClassName }}
   accessModes: 
     {{- toYaml .Values.inferenceCapability.pvc.accessModes | nindent 4 }}
   resources:

deployment/helm/charts/onyx/templates/inference-model-service.yaml

@@ -5,11 +5,11 @@ metadata:
 spec:
   type: {{ .Values.inferenceCapability.service.type }}
   ports:
-    - port: {{ .Values.inferenceCapability.service.servicePort}}
-      targetPort: {{ .Values.inferenceCapability.service.targetPort }}
+    - port: {{ .Values.inferenceCapability.service.port }}
+      targetPort: {{ .Values.inferenceCapability.service.port }}
       protocol: TCP
-      name: {{ .Values.inferenceCapability.service.portName }}
+      name: {{ .Values.inferenceCapability.service.name }}
   selector:
-    {{- range .Values.inferenceCapability.labels }}
+    {{- range .Values.inferenceCapability.deployment.labels }}
     {{ .key }}: {{ .value }}
     {{- end }}

deployment/helm/charts/onyx/templates/nginx-conf.yaml

@@ -5,11 +5,11 @@ metadata:
 data:
   nginx.conf: |
     upstream api_server {
-        server {{ include "onyx-stack.fullname" . }}-api-service:{{ .Values.api.service.servicePort }} fail_timeout=0;
+        server {{ include "onyx-stack.fullname" . }}-api-service:{{ .Values.api.service.port }} fail_timeout=0;
     }
 
     upstream web_server {
-        server {{ include "onyx-stack.fullname" . }}-webserver:{{ .Values.webserver.service.servicePort }} fail_timeout=0;
+        server {{ include "onyx-stack.fullname" . }}-webserver:{{ .Values.webserver.service.port }} fail_timeout=0;
     }
 
     server {

deployment/helm/charts/onyx/templates/tests/test-connection.yaml

@@ -11,5 +11,5 @@ spec:
     - name: wget
       image: busybox
       command: ['wget']
-      args: ['{{ include "onyx-stack.fullname" . }}-webserver:{{ .Values.webserver.service.servicePort }}']
+      args: ['{{ include "onyx-stack.fullname" . }}-webserver:{{ .Values.webserver.service.port }}']
   restartPolicy: Never

deployment/helm/charts/onyx/templates/webserver-deployment.yaml

@@ -41,7 +41,7 @@ spec:
           imagePullPolicy: {{ .Values.webserver.image.pullPolicy }}
           ports:
             - name: http
-              containerPort: {{ .Values.webserver.containerPorts.server }}
+              containerPort: {{ .Values.webserver.service.port }}
               protocol: TCP
           resources:
             {{- toYaml .Values.webserver.resources | nindent 12 }}

deployment/helm/charts/onyx/templates/webserver-service.yaml

@@ -10,8 +10,8 @@ metadata:
 spec:
   type: {{ .Values.webserver.service.type }}
   ports:
-    - port: {{ .Values.webserver.service.servicePort }}
-      targetPort: {{ .Values.webserver.service.targetPort }}
+    - port: {{ .Values.webserver.service.port }}
+      targetPort: http
       protocol: TCP
       name: http
   selector:

deployment/helm/charts/onyx/values.yaml

@@ -2,73 +2,62 @@
 # This is a YAML-formatted file.
 # Declare variables to be passed into your templates.
 
-postgresql:
-  primary:
-    persistence:
-      size: 5Gi
-  enabled: true
-  auth:
-    existingSecret: onyx-secrets
-    secretKeys:
-      # overwriting as postgres typically expects 'postgres-password'
-      adminPasswordKey: postgres_password
 imagePullSecrets: []
 nameOverride: ""
 fullnameOverride: ""
 
-persistent:
-  storageClassName: ""
-
 inferenceCapability:
   service:
-    portName: modelserver
+    name: inference-model-server-service
     type: ClusterIP
-    servicePort: 9000
-    targetPort: 9000
+    port: 9000
   pvc:
     name: inference-model-pvc
     accessModes:
       - ReadWriteOnce
     storage: 3Gi
-  name: inference-model-server
-  replicaCount: 1
-  labels:
-    - key: app
-      value: inference-model-server
-  image:
-    repository: onyxdotapp/onyx-model-server
-    # Overrides the image tag whose default is the chart appVersion.
-    tag: ""
-    pullPolicy: IfNotPresent
-  containerPorts:
-    server: 9000
-  volumeMounts:
-    - name: inference-model-storage
-      mountPath: /root/.cache
-  volumes:
-    - name: inference-model-storage
-      persistentVolumeClaim:
-        claimName: inference-model-pvc
+  deployment:
+    name: inference-model-server-deployment
+    replicas: 1
+    labels:
+      - key: app
+        value: inference-model-server
+    image:
+      repository: onyxdotapp/onyx-model-server
+      tag: latest
+      pullPolicy: IfNotPresent
+    command:
+      [
+        "uvicorn",
+        "model_server.main:app",
+        "--host",
+        "0.0.0.0",
+        "--port",
+        "9000",
+      ]
+    port: 9000
+    volumeMounts:
+      - name: inference-model-storage
+        mountPath: /root/.cache
+    volumes:
+      - name: inference-model-storage
+        persistentVolumeClaim:
+          claimName: inference-model-pvc
   podLabels:
     - key: app
       value: inference-model-server
 
 indexCapability:
   service:
-    portName: modelserver
     type: ClusterIP
-    servicePort: 9000
-    targetPort: 9000
-  replicaCount: 1
-  name: indexing-model-server
+    port: 9000
+    name: indexing-model-server-port
   deploymentLabels:
     app: indexing-model-server
   podLabels:
     app: indexing-model-server
   indexingOnly: "True"
   podAnnotations: {}
-  containerPorts:
-    server: 9000
   volumeMounts:
     - name: indexing-model-storage
       mountPath: /root/.cache
@@ -80,12 +69,7 @@ indexCapability:
     name: indexing-model-storage
     accessMode: "ReadWriteOnce"
     storage: "3Gi"
-  image:
-    repository: onyxdotapp/onyx-model-server
-    # Overrides the image tag whose default is the chart appVersion.
-    tag: ""
-    pullPolicy: IfNotPresent
-  limitConcurrency: 10
+
 config:
   envConfigMapName: env-configmap
 
@@ -100,6 +84,16 @@ serviceAccount:
   # If not set and create is true, a name is generated using the fullname template
   name: ""
 
+postgresql:
+  primary:
+    persistence:
+      size: 5Gi
+  enabled: true
+  auth:
+    existingSecret: onyx-secrets
+    secretKeys:
+      adminPasswordKey: postgres_password # overwriting as postgres typically expects 'postgres-password'
+
 nginx:
   containerPorts:
     http: 1024
@@ -141,13 +135,9 @@ webserver:
     # runAsNonRoot: true
     # runAsUser: 1000
 
-  containerPorts:
-    server: 3000
-
   service:
     type: ClusterIP
-    servicePort: 3000
-    targetPort: http
+    port: 3000
 
   resources: {}
   # We usually recommend not to specify default resources and to leave this as a conscious
@@ -166,7 +156,7 @@ webserver:
     minReplicas: 1
     maxReplicas: 100
     targetCPUUtilizationPercentage: 80
-    targetMemoryUtilizationPercentage: 80
+    # targetMemoryUtilizationPercentage: 80
 
   # Additional volumes on the output Deployment definition.
   volumes: []
@@ -199,9 +189,6 @@ api:
     scope: onyx-backend
     app: api-server
 
-  containerPorts:
-    server: 8080
-
   podSecurityContext:
     {}
     # fsGroup: 2000
@@ -217,9 +204,7 @@ api:
 
   service:
     type: ClusterIP
-    servicePort: 8080
-    targetPort: api-server-port
-    portName: api-server-port
+    port: 8080
 
   resources: {}
   # We usually recommend not to specify default resources and to leave this as a conscious
@@ -238,7 +223,7 @@ api:
     minReplicas: 1
     maxReplicas: 100
     targetCPUUtilizationPercentage: 80
-    targetMemoryUtilizationPercentage: 80
+    # targetMemoryUtilizationPercentage: 80
 
   # Additional volumes on the output Deployment definition.
   volumes: []
@@ -262,7 +247,7 @@ background:
     repository: onyxdotapp/onyx-backend
     pullPolicy: IfNotPresent
     # Overrides the image tag whose default is the chart appVersion.
-    tag: ""
+    tag: latest
   podAnnotations: {}
   podLabels:
     scope: onyx-backend
@@ -299,7 +284,7 @@ background:
     minReplicas: 1
     maxReplicas: 100
     targetCPUUtilizationPercentage: 80
-    targetMemoryUtilizationPercentage: 80
+    # targetMemoryUtilizationPercentage: 80
 
   # Additional volumes on the output Deployment definition.
   volumes: []
@@ -318,16 +303,6 @@ background:
   tolerations: []
 
 vespa:
-  volumeClaimTemplates:
-    - metadata:
-        name: vespa-storage
-      spec:
-        accessModes:
-          - ReadWriteOnce
-        storageClassName: ""
-        resources:
-          requests:
-            storage: 1Gi
   enabled: true
   replicaCount: 1
   image:
@@ -402,11 +377,19 @@ redis:
 #  #    hosts:
 #  #      - chart-example.local
 
+persistence:
+  vespa:
+    enabled: true
+    existingClaim: ""
+    storageClassName: ""
+    accessModes:
+      - ReadWriteOnce
+    size: 5Gi
 
 auth:
-  # existingSecret onyx-secret for storing smtp, oauth, slack, and other secrets
+  # for storing smtp, oauth, slack, and other secrets
   # keys are lowercased version of env vars (e.g. SMTP_USER -> smtp_user)
-  existingSecret: ""
+  existingSecret: "" # onyx-secrets
   # optionally override the secret keys to reference in the secret
   # this is used to populate the env vars in individual deployments
   # the values here reference the keys in secrets below
@@ -430,22 +413,14 @@ auth:
     redis_password: "password"
 
 configMap:
-  # Change this for production uses unless Onyx is only accessible behind VPN
-  AUTH_TYPE: "disabled"
-  # 1 Day Default
-  SESSION_EXPIRE_TIME_SECONDS: "86400"
-  # Can be something like onyx.app, as an extra double-check
-  VALID_EMAIL_DOMAINS: ""
-  # For sending verification emails, if unspecified then defaults to 'smtp.gmail.com'
-  SMTP_SERVER: ""
-  # For sending verification emails, if unspecified then defaults to '587'
-  SMTP_PORT: ""
-# 'your-email@company.com'
-  SMTP_USER: ""
-  # 'your-gmail-password'
-  # SMTP_PASS: ""
-  # 'your-email@company.com' SMTP_USER missing used instead
-  EMAIL_FROM: ""
+  AUTH_TYPE: "disabled" # Change this for production uses unless Onyx is only accessible behind VPN
+  SESSION_EXPIRE_TIME_SECONDS: "86400" # 1 Day Default
+  VALID_EMAIL_DOMAINS: "" # Can be something like onyx.app, as an extra double-check
+  SMTP_SERVER: "" # For sending verification emails, if unspecified then defaults to 'smtp.gmail.com'
+  SMTP_PORT: "" # For sending verification emails, if unspecified then defaults to '587'
+  SMTP_USER: "" # 'your-email@company.com'
+  # SMTP_PASS: ""  # 'your-gmail-password'
+  EMAIL_FROM: "" # 'your-email@company.com' SMTP_USER missing used instead
   # Gen AI Settings
   GEN_AI_MAX_TOKENS: ""
   QA_TIMEOUT: "60"
@@ -487,7 +462,7 @@ configMap:
   DANSWER_BOT_DISABLE_DOCS_ONLY_ANSWER: ""
   DANSWER_BOT_DISPLAY_ERROR_MSGS: ""
   DANSWER_BOT_RESPOND_EVERY_CHANNEL: ""
-  DANSWER_BOT_DISABLE_COT: ""
+  DANSWER_BOT_DISABLE_COT: "" # Currently unused
   NOTIFY_SLACKBOT_NO_ANSWER: ""
   # Logging
   # Optional Telemetry, please keep it on (nothing sensitive is collected)? <3
@@ -498,8 +473,7 @@ configMap:
   LOG_DANSWER_MODEL_INTERACTIONS: ""
   LOG_VESPA_TIMING_INFORMATION: ""
   # Shared or Non-backend Related
-  WEB_DOMAIN: "http://localhost:3000"
-  # DOMAIN used by nginx
-  DOMAIN: "localhost"
+  WEB_DOMAIN: "http://localhost:3000" # for web server and api server
+  DOMAIN: "localhost" # for nginx
   # Chat Configs
   HARD_DELETE_CHATS: ""

web/package-lock.json

@@ -25,7 +25,6 @@
         "@radix-ui/react-scroll-area": "^1.2.2",
         "@radix-ui/react-select": "^2.1.2",
         "@radix-ui/react-separator": "^1.1.0",
-        "@radix-ui/react-slider": "^1.2.2",
         "@radix-ui/react-slot": "^1.1.0",
         "@radix-ui/react-switch": "^1.1.1",
         "@radix-ui/react-tabs": "^1.1.1",
@@ -4964,142 +4963,6 @@
         }
       }
     },
-    "node_modules/@radix-ui/react-slider": {
-      "version": "1.2.2",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-slider/-/react-slider-1.2.2.tgz",
-      "integrity": "sha512-sNlU06ii1/ZcbHf8I9En54ZPW0Vil/yPVg4vQMcFNjrIx51jsHbFl1HYHQvCIWJSr1q0ZmA+iIs/ZTv8h7HHSA==",
-      "license": "MIT",
-      "dependencies": {
-        "@radix-ui/number": "1.1.0",
-        "@radix-ui/primitive": "1.1.1",
-        "@radix-ui/react-collection": "1.1.1",
-        "@radix-ui/react-compose-refs": "1.1.1",
-        "@radix-ui/react-context": "1.1.1",
-        "@radix-ui/react-direction": "1.1.0",
-        "@radix-ui/react-primitive": "2.0.1",
-        "@radix-ui/react-use-controllable-state": "1.1.0",
-        "@radix-ui/react-use-layout-effect": "1.1.0",
-        "@radix-ui/react-use-previous": "1.1.0",
-        "@radix-ui/react-use-size": "1.1.0"
-      },
-      "peerDependencies": {
-        "@types/react": "*",
-        "@types/react-dom": "*",
-        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
-        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
-      },
-      "peerDependenciesMeta": {
-        "@types/react": {
-          "optional": true
-        },
-        "@types/react-dom": {
-          "optional": true
-        }
-      }
-    },
-    "node_modules/@radix-ui/react-slider/node_modules/@radix-ui/primitive": {
-      "version": "1.1.1",
-      "resolved": "https://registry.npmjs.org/@radix-ui/primitive/-/primitive-1.1.1.tgz",
-      "integrity": "sha512-SJ31y+Q/zAyShtXJc8x83i9TYdbAfHZ++tUZnvjJJqFjzsdUnKsxPL6IEtBlxKkU7yzer//GQtZSV4GbldL3YA==",
-      "license": "MIT"
-    },
-    "node_modules/@radix-ui/react-slider/node_modules/@radix-ui/react-collection": {
-      "version": "1.1.1",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-collection/-/react-collection-1.1.1.tgz",
-      "integrity": "sha512-LwT3pSho9Dljg+wY2KN2mrrh6y3qELfftINERIzBUO9e0N+t0oMTyn3k9iv+ZqgrwGkRnLpNJrsMv9BZlt2yuA==",
-      "license": "MIT",
-      "dependencies": {
-        "@radix-ui/react-compose-refs": "1.1.1",
-        "@radix-ui/react-context": "1.1.1",
-        "@radix-ui/react-primitive": "2.0.1",
-        "@radix-ui/react-slot": "1.1.1"
-      },
-      "peerDependencies": {
-        "@types/react": "*",
-        "@types/react-dom": "*",
-        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
-        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
-      },
-      "peerDependenciesMeta": {
-        "@types/react": {
-          "optional": true
-        },
-        "@types/react-dom": {
-          "optional": true
-        }
-      }
-    },
-    "node_modules/@radix-ui/react-slider/node_modules/@radix-ui/react-compose-refs": {
-      "version": "1.1.1",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-compose-refs/-/react-compose-refs-1.1.1.tgz",
-      "integrity": "sha512-Y9VzoRDSJtgFMUCoiZBDVo084VQ5hfpXxVE+NgkdNsjiDBByiImMZKKhxMwCbdHvhlENG6a833CbFkOQvTricw==",
-      "license": "MIT",
-      "peerDependencies": {
-        "@types/react": "*",
-        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
-      },
-      "peerDependenciesMeta": {
-        "@types/react": {
-          "optional": true
-        }
-      }
-    },
-    "node_modules/@radix-ui/react-slider/node_modules/@radix-ui/react-context": {
-      "version": "1.1.1",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-context/-/react-context-1.1.1.tgz",
-      "integrity": "sha512-UASk9zi+crv9WteK/NU4PLvOoL3OuE6BWVKNF6hPRBtYBDXQ2u5iu3O59zUlJiTVvkyuycnqrztsHVJwcK9K+Q==",
-      "license": "MIT",
-      "peerDependencies": {
-        "@types/react": "*",
-        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
-      },
-      "peerDependenciesMeta": {
-        "@types/react": {
-          "optional": true
-        }
-      }
-    },
-    "node_modules/@radix-ui/react-slider/node_modules/@radix-ui/react-primitive": {
-      "version": "2.0.1",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.0.1.tgz",
-      "integrity": "sha512-sHCWTtxwNn3L3fH8qAfnF3WbUZycW93SM1j3NFDzXBiz8D6F5UTTy8G1+WFEaiCdvCVRJWj6N2R4Xq6HdiHmDg==",
-      "license": "MIT",
-      "dependencies": {
-        "@radix-ui/react-slot": "1.1.1"
-      },
-      "peerDependencies": {
-        "@types/react": "*",
-        "@types/react-dom": "*",
-        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
-        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
-      },
-      "peerDependenciesMeta": {
-        "@types/react": {
-          "optional": true
-        },
-        "@types/react-dom": {
-          "optional": true
-        }
-      }
-    },
-    "node_modules/@radix-ui/react-slider/node_modules/@radix-ui/react-slot": {
-      "version": "1.1.1",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-slot/-/react-slot-1.1.1.tgz",
-      "integrity": "sha512-RApLLOcINYJA+dMVbOju7MYv1Mb2EBp2nH4HdDzXTSyaR5optlm6Otrz1euW3HbdOR8UmmFK06TD+A9frYWv+g==",
-      "license": "MIT",
-      "dependencies": {
-        "@radix-ui/react-compose-refs": "1.1.1"
-      },
-      "peerDependencies": {
-        "@types/react": "*",
-        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
-      },
-      "peerDependenciesMeta": {
-        "@types/react": {
-          "optional": true
-        }
-      }
-    },
     "node_modules/@radix-ui/react-slot": {
       "version": "1.1.0",
       "resolved": "https://registry.npmjs.org/@radix-ui/react-slot/-/react-slot-1.1.0.tgz",

web/package.json

@@ -28,7 +28,6 @@
     "@radix-ui/react-scroll-area": "^1.2.2",
     "@radix-ui/react-select": "^2.1.2",
     "@radix-ui/react-separator": "^1.1.0",
-    "@radix-ui/react-slider": "^1.2.2",
     "@radix-ui/react-slot": "^1.1.0",
     "@radix-ui/react-switch": "^1.1.1",
     "@radix-ui/react-tabs": "^1.1.1",

web/playwright.config.ts

@@ -1,17 +1,41 @@
 import { defineConfig, devices } from "@playwright/test";
 
 export default defineConfig({
-  globalSetup: require.resolve("./tests/e2e/global-setup"),
-  timeout: 30000, // 30 seconds timeout
+  workers: 1, // temporary change to see if single threaded testing stabilizes the tests
+  testDir: "./tests/e2e", // Folder for test files
+  reporter: "list",
+  // Configure paths for screenshots
+  // expect: {
+  //   toMatchSnapshot: {
+  //     threshold: 0.2, // Adjust the threshold for visual diffs
+  //   },
+  // },
+  // reporter: [["html", { outputFolder: "test-results/output/report" }]], // HTML report location
+  // outputDir: "test-results/output/screenshots", // Set output folder for test artifacts
   projects: [
     {
-      name: "admin",
+      // dependency for admin workflows
+      name: "admin_setup",
+      testMatch: /.*\admin_auth\.setup\.ts/,
+    },
+    {
+      // tests admin workflows
+      name: "chromium-admin",
+      grep: /@admin/,
       use: {
         ...devices["Desktop Chrome"],
-        viewport: { width: 1280, height: 720 },
+        // Use prepared auth state.
         storageState: "admin_auth.json",
       },
-      testIgnore: ["**/codeUtils.test.ts"],
+      dependencies: ["admin_setup"],
+    },
+    {
+      // tests logged out / guest workflows
+      name: "chromium-guest",
+      grep: /@guest/,
+      use: {
+        ...devices["Desktop Chrome"],
+      },
     },
   ],
 });

web/src/app/admin/assistants/AssistantEditor.tsx

@@ -720,6 +720,7 @@ export function AssistantEditor({
                 name="description"
                 label="Description"
                 placeholder="Use this Assistant to help draft professional emails"
+                data-testid="assistant-description-input"
                 className="[&_input]:placeholder:text-text-muted/50"
               />
 

web/src/app/chat/ChatIntro.tsx

@@ -4,7 +4,7 @@ import { OnyxIcon } from "@/components/icons/icons";
 
 export function ChatIntro({ selectedPersona }: { selectedPersona: Persona }) {
   return (
-    <div data-testid="chat-intro" className="flex flex-col items-center gap-6">
+    <div className="flex flex-col items-center gap-6">
       <div className="relative flex flex-col gap-y-4 w-fit mx-auto justify-center">
         <div className="absolute z-10 items-center flex -left-12 top-1/2 -translate-y-1/2">
           <AssistantIcon size={36} assistant={selectedPersona} />

web/src/app/chat/ChatPage.tsx

@@ -111,7 +111,6 @@ import {
 import AssistantModal from "../assistants/mine/AssistantModal";
 import { getSourceMetadata } from "@/lib/sources";
 import { UserSettingsModal } from "./modal/UserSettingsModal";
-import { AlignStartVertical } from "lucide-react";
 
 const TEMP_USER_MESSAGE_ID = -1;
 const TEMP_ASSISTANT_MESSAGE_ID = -2;
@@ -190,11 +189,7 @@ export function ChatPage({
 
   const [userSettingsToggled, setUserSettingsToggled] = useState(false);
 
-  const {
-    assistants: availableAssistants,
-    finalAssistants,
-    pinnedAssistants,
-  } = useAssistants();
+  const { assistants: availableAssistants, finalAssistants } = useAssistants();
 
   const [showApiKeyModal, setShowApiKeyModal] = useState(
     !shouldShowWelcomeModal
@@ -277,6 +272,16 @@ export function ChatPage({
     SEARCH_PARAM_NAMES.TEMPERATURE
   );
 
+  const defaultTemperature = search_param_temperature
+    ? parseFloat(search_param_temperature)
+    : selectedAssistant?.tools.some(
+          (tool) =>
+            tool.in_code_tool_id === SEARCH_TOOL_ID ||
+            tool.in_code_tool_id === INTERNET_SEARCH_TOOL_ID
+        )
+      ? 0
+      : 0.7;
+
   const setSelectedAssistantFromId = (assistantId: number) => {
     // NOTE: also intentionally look through available assistants here, so that
     // even if the user has hidden an assistant they can still go back to it
@@ -292,22 +297,20 @@ export function ChatPage({
   const [presentingDocument, setPresentingDocument] =
     useState<OnyxDocument | null>(null);
 
-  // Current assistant is decided based on this ordering
-  // 1. Alternative assistant (assistant selected explicitly by user)
-  // 2. Selected assistant (assistnat default in this chat session)
-  // 3. First pinned assistants (ordered list of pinned assistants)
-  // 4. Available assistants (ordered list of available assistants)
-  // Relevant test: `live_assistant.spec.ts`
+  const { recentAssistants, refreshRecentAssistants } = useAssistants();
+
   const liveAssistant: Persona | undefined = useMemo(
     () =>
       alternativeAssistant ||
       selectedAssistant ||
-      pinnedAssistants[0] ||
+      recentAssistants[0] ||
+      finalAssistants[0] ||
       availableAssistants[0],
     [
       alternativeAssistant,
       selectedAssistant,
-      pinnedAssistants,
+      recentAssistants,
+      finalAssistants,
       availableAssistants,
     ]
   );
@@ -404,6 +407,9 @@ export function ChatPage({
       filterManager.setSelectedTags([]);
       filterManager.setTimeRange(null);
 
+      // reset LLM overrides (based on chat session!)
+      llmOverrideManager.updateTemperature(null);
+
       // remove uploaded files
       setCurrentMessageFiles([]);
 
@@ -446,7 +452,6 @@ export function ChatPage({
       );
 
       const chatSession = (await response.json()) as BackendChatSession;
-
       setSelectedAssistantFromId(chatSession.persona_id);
 
       const newMessageMap = processRawChatHistory(chatSession.messages);
@@ -811,6 +816,7 @@ export function ChatPage({
         setMaxTokens(maxTokens);
       }
     }
+    refreshRecentAssistants(liveAssistant?.id);
     fetchMaxTokens();
   }, [liveAssistant]);
 

web/src/app/chat/input/ChatInputBar.tsx

@@ -478,7 +478,6 @@ export function ChatInputBar({
               onKeyDownCapture={handleKeyDown}
               onChange={handleInputChange}
               ref={textAreaRef}
-              id="onyx-chat-input-textarea"
               className={`
                 m-0
                 w-full
@@ -704,7 +703,6 @@ export function ChatInputBar({
               </div>
               <div className="flex my-auto">
                 <button
-                  id="onyx-chat-input-send-button"
                   className={`cursor-pointer ${
                     chatState == "streaming" ||
                     chatState == "toolBuilding" ||

web/src/app/chat/input/LLMPopover.tsx

@@ -1,4 +1,4 @@
-import React, { useState, useEffect } from "react";
+import React, { useState } from "react";
 import {
   Popover,
   PopoverContent,
@@ -26,9 +26,6 @@ import {
 } from "@/components/ui/tooltip";
 import { FiAlertTriangle } from "react-icons/fi";
 
-import { Slider } from "@/components/ui/slider";
-import { useUser } from "@/components/user/UserProvider";
-
 interface LLMPopoverProps {
   llmProviders: LLMProviderDescriptor[];
   llmOverrideManager: LlmOverrideManager;
@@ -43,7 +40,6 @@ export default function LLMPopover({
   currentAssistant,
 }: LLMPopoverProps) {
   const [isOpen, setIsOpen] = useState(false);
-  const { user } = useUser();
   const { llmOverride, updateLLMOverride } = llmOverrideManager;
   const currentLlm = llmOverride.modelName;
 
@@ -92,29 +88,10 @@ export default function LLMPopover({
     ? getDisplayNameForModel(defaultModelName)
     : null;
 
-  const [localTemperature, setLocalTemperature] = useState(
-    llmOverrideManager.temperature ?? 0.5
-  );
-
-  useEffect(() => {
-    setLocalTemperature(llmOverrideManager.temperature ?? 0.5);
-  }, [llmOverrideManager.temperature]);
-
-  const handleTemperatureChange = (value: number[]) => {
-    setLocalTemperature(value[0]);
-  };
-
-  const handleTemperatureChangeComplete = (value: number[]) => {
-    llmOverrideManager.updateTemperature(value[0]);
-  };
-
   return (
     <Popover open={isOpen} onOpenChange={setIsOpen}>
       <PopoverTrigger asChild>
-        <button
-          className="focus:outline-none"
-          data-testid="llm-popover-trigger"
-        >
+        <button className="focus:outline-none">
           <ChatInputOption
             minimize
             toggle
@@ -138,9 +115,9 @@ export default function LLMPopover({
       </PopoverTrigger>
       <PopoverContent
         align="start"
-        className="w-64 p-1 bg-background border border-gray-200 rounded-md shadow-lg flex flex-col"
+        className="w-64 p-1 bg-background border border-gray-200 rounded-md shadow-lg"
       >
-        <div className="flex-grow max-h-[300px] default-scrollbar overflow-y-auto">
+        <div className="max-h-[300px] overflow-y-auto">
           {llmOptions.map(({ name, icon, value }, index) => {
             if (!requiresImageGeneration || checkLLMSupportsImageInput(name)) {
               return (
@@ -191,25 +168,6 @@ export default function LLMPopover({
             return null;
           })}
         </div>
-        {user?.preferences?.temperature_override_enabled && (
-          <div className="mt-2 pt-2 border-t border-gray-200">
-            <div className="w-full px-3 py-2">
-              <Slider
-                value={[localTemperature]}
-                max={llmOverrideManager.maxTemperature}
-                min={0}
-                step={0.01}
-                onValueChange={handleTemperatureChange}
-                onValueCommit={handleTemperatureChangeComplete}
-                className="w-full"
-              />
-              <div className="flex justify-between text-xs text-gray-500 mt-2">
-                <span>Temperature (creativity)</span>
-                <span>{localTemperature.toFixed(1)}</span>
-              </div>
-            </div>
-          </div>
-        )}
       </PopoverContent>
     </Popover>
   );

web/src/app/chat/interfaces.ts

@@ -68,7 +68,6 @@ export interface ChatSession {
   shared_status: ChatSessionSharedStatus;
   folder_id: number | null;
   current_alternate_model: string;
-  current_temperature_override: number | null;
 }
 
 export interface SearchSession {
@@ -108,7 +107,6 @@ export interface BackendChatSession {
   messages: BackendMessage[];
   time_created: string;
   shared_status: ChatSessionSharedStatus;
-  current_temperature_override: number | null;
   current_alternate_model?: string;
 }
 

web/src/app/chat/lib.tsx

@@ -75,23 +75,6 @@ export async function updateModelOverrideForChatSession(
   return response;
 }
 
-export async function updateTemperatureOverrideForChatSession(
-  chatSessionId: string,
-  newTemperature: number
-) {
-  const response = await fetch("/api/chat/update-chat-session-temperature", {
-    method: "PUT",
-    headers: {
-      "Content-Type": "application/json",
-    },
-    body: JSON.stringify({
-      chat_session_id: chatSessionId,
-      temperature_override: newTemperature,
-    }),
-  });
-  return response;
-}
-
 export async function createChatSession(
   personaId: number,
   description: string | null

web/src/app/chat/message/Messages.tsx

@@ -402,7 +402,7 @@ export const AIMessage = ({
 
   return (
     <div
-      id={isComplete ? "onyx-ai-message" : undefined}
+      id="onyx-ai-message"
       ref={trackedElementRef}
       className={`py-5 ml-4 lg:px-5 relative flex `}
     >

web/src/app/chat/modal/UserSettingsModal.tsx

@@ -30,13 +30,8 @@ export function UserSettingsModal({
   defaultModel: string | null;
 }) {
   const { inputPrompts, refreshInputPrompts } = useChatContext();
-  const {
-    refreshUser,
-    user,
-    updateUserAutoScroll,
-    updateUserShortcuts,
-    updateUserTemperatureOverrideEnabled,
-  } = useUser();
+  const { refreshUser, user, updateUserAutoScroll, updateUserShortcuts } =
+    useUser();
   const containerRef = useRef<HTMLDivElement>(null);
   const messageRef = useRef<HTMLDivElement>(null);
 
@@ -184,16 +179,6 @@ export function UserSettingsModal({
             />
             <Label className="text-sm">Enable Prompt Shortcuts</Label>
           </div>
-          <div className="flex items-center gap-x-2">
-            <Switch
-              size="sm"
-              checked={user?.preferences?.temperature_override_enabled}
-              onCheckedChange={(checked) => {
-                updateUserTemperatureOverrideEnabled(checked);
-              }}
-            />
-            <Label className="text-sm">Enable Temperature Override</Label>
-          </div>
         </div>
 
         <Separator />

web/src/app/chat/sessionSidebar/HistorySidebar.tsx

@@ -19,7 +19,9 @@ import {
 
 import { useRouter, useSearchParams } from "next/navigation";
 import { ChatSession } from "../interfaces";
+import { NEXT_PUBLIC_NEW_CHAT_DIRECTS_TO_SAME_PERSONA } from "@/lib/constants";
 import { Folder } from "../folders/interfaces";
+import { usePopup } from "@/components/admin/connectors/Popup";
 import { SettingsContext } from "@/components/settings/SettingsProvider";
 
 import { DocumentIcon2, NewChatIcon } from "@/components/icons/icons";
@@ -129,7 +131,6 @@ const SortableAssistant: React.FC<SortableAssistantProps> = ({
         className="w-3 ml-[2px] mr-[2px] group-hover:visible invisible flex-none cursor-grab"
       />
       <button
-        data-testid={`assistant-[${assistant.id}]`}
         onClick={(e) => {
           e.preventDefault();
           if (!isDragging) {
@@ -250,11 +251,9 @@ export const HistorySidebar = forwardRef<HTMLDivElement, HistorySidebarProps>(
 
     const handleNewChat = () => {
       reset();
-      console.log("currentChatSession", currentChatSession);
-
       const newChatUrl =
         `/${page}` +
-        (currentChatSession
+        (NEXT_PUBLIC_NEW_CHAT_DIRECTS_TO_SAME_PERSONA && currentChatSession
           ? `?assistantId=${currentChatSession.persona_id}`
           : "");
       router.push(newChatUrl);
@@ -276,6 +275,7 @@ export const HistorySidebar = forwardRef<HTMLDivElement, HistorySidebarProps>(
             flex-col relative
             h-screen
             pt-2
+            
             transition-transform 
             `}
         >
@@ -294,7 +294,8 @@ export const HistorySidebar = forwardRef<HTMLDivElement, HistorySidebarProps>(
                 className="w-full px-2 py-1  rounded-md items-center hover:bg-hover cursor-pointer transition-all duration-150 flex gap-x-2"
                 href={
                   `/${page}` +
-                  (currentChatSession
+                  (NEXT_PUBLIC_NEW_CHAT_DIRECTS_TO_SAME_PERSONA &&
+                  currentChatSession?.persona_id
                     ? `?assistantId=${currentChatSession?.persona_id}`
                     : "")
                 }
@@ -319,6 +320,14 @@ export const HistorySidebar = forwardRef<HTMLDivElement, HistorySidebarProps>(
                 <Link
                   className="w-full px-2 py-1  rounded-md items-center hover:bg-hover cursor-pointer transition-all duration-150 flex gap-x-2"
                   href="/chat/input-prompts"
+                  onClick={(e) => {
+                    if (e.metaKey || e.ctrlKey) {
+                      return;
+                    }
+                    if (handleNewChat) {
+                      handleNewChat();
+                    }
+                  }}
                 >
                   <DocumentIcon2
                     size={20}

web/src/components/Modal.tsx

@@ -103,7 +103,7 @@ export function Modal({
             </button>
           </div>
         )}
-        <div className="items-start flex-shrink-0">
+        <div className="flex-shrink-0">
           {title && (
             <>
               <div className="flex">

web/src/components/UserDropdown.tsx

@@ -133,7 +133,6 @@ export function UserDropdown({
         onOpenChange={onOpenChange}
         content={
           <div
-            id="onyx-user-dropdown"
             onClick={() => setUserInfoVisible(!userInfoVisible)}
             className="flex relative cursor-pointer"
           >

web/src/components/admin/connectors/Popup.tsx

@@ -59,23 +59,7 @@ export const Popup: React.FC<PopupSpec> = ({ message, type }) => (
         />
       </svg>
     )}
-    <div className="flex flex-col items-center space-x-2">
-      <span className="font-medium">{message}</span>
-      {type === "error" && (
-        <span className="text-xs text-red-100">
-          Need help?{" "}
-          <a
-            href="https://join.slack.com/t/onyx-dot-app/shared_invite/zt-2twesxdr6-5iQitKZQpgq~hYIZ~dv3KA"
-            target="_blank"
-            rel="noopener noreferrer"
-            className="underline hover:text-red-900"
-          >
-            Join our community
-          </a>{" "}
-          for support.
-        </span>
-      )}
-    </div>
+    <span className="font-medium">{message}</span>
   </div>
 );
 

web/src/components/chat_search/Header.tsx

@@ -2,6 +2,7 @@
 import { UserDropdown } from "../UserDropdown";
 import { FiShare2 } from "react-icons/fi";
 import { SetStateAction, useContext, useEffect } from "react";
+import { NEXT_PUBLIC_NEW_CHAT_DIRECTS_TO_SAME_PERSONA } from "@/lib/constants";
 import { ChatSession } from "@/app/chat/interfaces";
 import Link from "next/link";
 import { pageType } from "@/app/chat/sessionSidebar/types";
@@ -41,7 +42,8 @@ export default function FunctionalHeader({
             event.preventDefault();
             window.open(
               `/${page}` +
-                (currentChatSession
+                (NEXT_PUBLIC_NEW_CHAT_DIRECTS_TO_SAME_PERSONA &&
+                currentChatSession
                   ? `?assistantId=${currentChatSession.persona_id}`
                   : ""),
               "_self"
@@ -61,7 +63,7 @@ export default function FunctionalHeader({
     reset();
     const newChatUrl =
       `/${page}` +
-      (currentChatSession
+      (NEXT_PUBLIC_NEW_CHAT_DIRECTS_TO_SAME_PERSONA && currentChatSession
         ? `?assistantId=${currentChatSession.persona_id}`
         : "");
     router.push(newChatUrl);
@@ -126,6 +128,25 @@ export default function FunctionalHeader({
             </div>
           )}
 
+          {/* <div
+            className={`absolute 
+              ${
+                documentSidebarToggled && !sidebarToggled
+                  ? "left-[calc(50%-125px)]"
+                  : !documentSidebarToggled && sidebarToggled
+                    ? "left-[calc(50%+125px)]"
+                    : "left-1/2"
+              }
+              ${
+                documentSidebarToggled || sidebarToggled
+                  ? "mobile:w-[40vw] max-w-[50vw]"
+                  : "mobile:w-[50vw] max-w-[60vw]"
+              }
+              top-1/2 transform -translate-x-1/2 -translate-y-1/2 transition-all duration-300`}
+          >
+            <ChatBanner />
+          </div> */}
+
           <div className="invisible">
             <LogoWithText
               page={page}
@@ -135,6 +156,8 @@ export default function FunctionalHeader({
             />
           </div>
 
+          {/* className="fixed cursor-pointer flex z-40 left-4 bg-black top-3 h-8" */}
+
           <div className="absolute right-2 mobile:top-1 desktop:top-1 h-8  flex">
             {setSharingModalVisible && !hideUserDropdown && (
               <div
@@ -156,7 +179,8 @@ export default function FunctionalHeader({
               className="desktop:hidden ml-2 my-auto"
               href={
                 `/${page}` +
-                (currentChatSession
+                (NEXT_PUBLIC_NEW_CHAT_DIRECTS_TO_SAME_PERSONA &&
+                currentChatSession
                   ? `?assistantId=${currentChatSession.persona_id}`
                   : "")
               }

web/src/components/context/AssistantsContext.tsx

@@ -25,6 +25,8 @@ interface AssistantsContextProps {
   ownedButHiddenAssistants: Persona[];
   refreshAssistants: () => Promise<void>;
   isImageGenerationAvailable: boolean;
+  recentAssistants: Persona[];
+  refreshRecentAssistants: (currentAssistant: number) => Promise<void>;
   // Admin only
   editablePersonas: Persona[];
   allAssistants: Persona[];
@@ -54,28 +56,35 @@ export const AssistantsProvider: React.FC<{
   const [editablePersonas, setEditablePersonas] = useState<Persona[]>([]);
   const [allAssistants, setAllAssistants] = useState<Persona[]>([]);
 
-  const [pinnedAssistants, setPinnedAssistants] = useState<Persona[]>(() => {
-    if (user?.preferences.pinned_assistants) {
-      return user.preferences.pinned_assistants
-        .map((id) => assistants.find((assistant) => assistant.id === id))
-        .filter((assistant): assistant is Persona => assistant !== undefined);
-    } else {
-      return assistants.filter((a) => a.builtin_persona);
-    }
-  });
+  const [pinnedAssistants, setPinnedAssistants] = useState<Persona[]>(
+    user?.preferences.pinned_assistants
+      ? assistants.filter((assistant) =>
+          user?.preferences?.pinned_assistants?.includes(assistant.id)
+        )
+      : assistants.filter((a) => a.builtin_persona)
+  );
 
   useEffect(() => {
-    setPinnedAssistants(() => {
-      if (user?.preferences.pinned_assistants) {
-        return user.preferences.pinned_assistants
-          .map((id) => assistants.find((assistant) => assistant.id === id))
-          .filter((assistant): assistant is Persona => assistant !== undefined);
-      } else {
-        return assistants.filter((a) => a.builtin_persona);
-      }
-    });
+    setPinnedAssistants(
+      user?.preferences.pinned_assistants
+        ? assistants.filter((assistant) =>
+            user?.preferences?.pinned_assistants?.includes(assistant.id)
+          )
+        : assistants.filter((a) => a.builtin_persona)
+    );
   }, [user?.preferences?.pinned_assistants, assistants]);
 
+  const [recentAssistants, setRecentAssistants] = useState<Persona[]>(
+    user?.preferences.recent_assistants
+      ?.filter((assistantId) =>
+        assistants.find((assistant) => assistant.id === assistantId)
+      )
+      .map(
+        (assistantId) =>
+          assistants.find((assistant) => assistant.id === assistantId)!
+      ) || []
+  );
+
   const [isImageGenerationAvailable, setIsImageGenerationAvailable] =
     useState<boolean>(false);
 
@@ -126,6 +135,28 @@ export const AssistantsProvider: React.FC<{
     fetchPersonas();
   }, [isAdmin, isCurator]);
 
+  const refreshRecentAssistants = async (currentAssistant: number) => {
+    const response = await fetch("/api/user/recent-assistants", {
+      method: "PATCH",
+      headers: {
+        "Content-Type": "application/json",
+      },
+      body: JSON.stringify({
+        current_assistant: currentAssistant,
+      }),
+    });
+    if (!response.ok) {
+      return;
+    }
+    setRecentAssistants((recentAssistants) => [
+      assistants.find((assistant) => assistant.id === currentAssistant)!,
+
+      ...recentAssistants.filter(
+        (assistant) => assistant.id !== currentAssistant
+      ),
+    ]);
+  };
+
   const refreshAssistants = async () => {
     try {
       const response = await fetch("/api/persona", {
@@ -150,6 +181,13 @@ export const AssistantsProvider: React.FC<{
     } catch (error) {
       console.error("Error refreshing assistants:", error);
     }
+
+    setRecentAssistants(
+      assistants.filter(
+        (assistant) =>
+          user?.preferences.recent_assistants?.includes(assistant.id) || false
+      )
+    );
   };
 
   const {
@@ -192,6 +230,8 @@ export const AssistantsProvider: React.FC<{
         editablePersonas,
         allAssistants,
         isImageGenerationAvailable,
+        recentAssistants,
+        refreshRecentAssistants,
         setPinnedAssistants,
         pinnedAssistants,
       }}

web/src/components/header/LogoWithText.tsx

@@ -2,6 +2,7 @@
 import { useContext } from "react";
 import { FiSidebar } from "react-icons/fi";
 import { SettingsContext } from "../settings/SettingsProvider";
+import { NEXT_PUBLIC_NEW_CHAT_DIRECTS_TO_SAME_PERSONA } from "@/lib/constants";
 import { LeftToLineIcon, NewChatIcon, RightToLineIcon } from "../icons/icons";
 import {
   Tooltip,
@@ -89,7 +90,9 @@ export default function LogoWithText({
                 className="my-auto mobile:hidden"
                 href={
                   `/${page}` +
-                  (assistantId ? `?assistantId=${assistantId}` : "")
+                  (NEXT_PUBLIC_NEW_CHAT_DIRECTS_TO_SAME_PERSONA && assistantId
+                    ? `?assistantId=${assistantId}`
+                    : "")
                 }
                 onClick={(e) => {
                   if (e.metaKey || e.ctrlKey) {

web/src/components/ui/slider.tsx

@@ -1,28 +0,0 @@
-"use client";
-
-import * as React from "react";
-import * as SliderPrimitive from "@radix-ui/react-slider";
-
-import { cn } from "@/lib/utils";
-
-const Slider = React.forwardRef<
-  React.ElementRef<typeof SliderPrimitive.Root>,
-  React.ComponentPropsWithoutRef<typeof SliderPrimitive.Root>
->(({ className, ...props }, ref) => (
-  <SliderPrimitive.Root
-    ref={ref}
-    className={cn(
-      "relative flex w-full touch-none select-none items-center",
-      className
-    )}
-    {...props}
-  >
-    <SliderPrimitive.Track className="relative h-2 w-full grow overflow-hidden rounded-full bg-neutral-100 dark:bg-neutral-800">
-      <SliderPrimitive.Range className="absolute h-full bg-neutral-900 dark:bg-neutral-50" />
-    </SliderPrimitive.Track>
-    <SliderPrimitive.Thumb className="block h-3 w-3 rounded-full border border-neutral-900 bg-white ring-offset-white transition-colors focus-visible:outline-none focus-visible:ring-2 focus-visible:ring-neutral-950 focus-visible:ring-offset disabled:pointer-events-none disabled:opacity-50 dark:border-neutral-50 dark:bg-neutral-950 dark:ring-offset-neutral-950 dark:focus-visible:ring-neutral-300" />
-  </SliderPrimitive.Root>
-));
-Slider.displayName = SliderPrimitive.Root.displayName;
-
-export { Slider };

web/src/components/user/UserProvider.tsx

@@ -18,7 +18,6 @@ interface UserContextType {
     assistantId: number,
     isPinned: boolean
   ) => Promise<boolean>;
-  updateUserTemperatureOverrideEnabled: (enabled: boolean) => Promise<void>;
 }
 
 const UserContext = createContext<UserContextType | undefined>(undefined);
@@ -58,41 +57,6 @@ export function UserProvider({
       console.error("Error fetching current user:", error);
     }
   };
-  const updateUserTemperatureOverrideEnabled = async (enabled: boolean) => {
-    try {
-      setUpToDateUser((prevUser) => {
-        if (prevUser) {
-          return {
-            ...prevUser,
-            preferences: {
-              ...prevUser.preferences,
-              temperature_override_enabled: enabled,
-            },
-          };
-        }
-        return prevUser;
-      });
-
-      const response = await fetch(
-        `/api/temperature-override-enabled?temperature_override_enabled=${enabled}`,
-        {
-          method: "PATCH",
-          headers: {
-            "Content-Type": "application/json",
-          },
-        }
-      );
-
-      if (!response.ok) {
-        await refreshUser();
-        throw new Error("Failed to update user temperature override setting");
-      }
-    } catch (error) {
-      console.error("Error updating user temperature override setting:", error);
-      throw error;
-    }
-  };
-
   const updateUserShortcuts = async (enabled: boolean) => {
     try {
       setUpToDateUser((prevUser) => {
@@ -220,7 +184,6 @@ export function UserProvider({
         refreshUser,
         updateUserAutoScroll,
         updateUserShortcuts,
-        updateUserTemperatureOverrideEnabled,
         toggleAssistantPinnedStatus,
         isAdmin: upToDateUser?.role === UserRole.ADMIN,
         // Curator status applies for either global or basic curator

web/src/lib/constants.ts

@@ -18,6 +18,10 @@ export const NEXT_PUBLIC_DO_NOT_USE_TOGGLE_OFF_DANSWER_POWERED =
   process.env.NEXT_PUBLIC_DO_NOT_USE_TOGGLE_OFF_DANSWER_POWERED?.toLowerCase() ===
   "true";
 
+export const NEXT_PUBLIC_NEW_CHAT_DIRECTS_TO_SAME_PERSONA =
+  process.env.NEXT_PUBLIC_NEW_CHAT_DIRECTS_TO_SAME_PERSONA?.toLowerCase() ===
+  "true";
+
 export const GMAIL_AUTH_IS_ADMIN_COOKIE_NAME = "gmail_auth_is_admin";
 
 export const GOOGLE_DRIVE_AUTH_IS_ADMIN_COOKIE_NAME =

web/src/lib/hooks.ts

@@ -10,7 +10,7 @@ import {
 } from "@/lib/types";
 import useSWR, { mutate, useSWRConfig } from "swr";
 import { errorHandlingFetcher } from "./fetcher";
-import { useContext, useEffect, useMemo, useState } from "react";
+import { useContext, useEffect, useState } from "react";
 import { DateRangePickerValue } from "@/app/ee/admin/performance/DateRangeSelector";
 import { Filters, SourceMetadata } from "./search/interfaces";
 import {
@@ -28,8 +28,6 @@ import { isAnthropic } from "@/app/admin/configuration/llm/interfaces";
 import { getSourceMetadata } from "./sources";
 import { AuthType, NEXT_PUBLIC_CLOUD_ENABLED } from "./constants";
 import { useUser } from "@/components/user/UserProvider";
-import { SEARCH_TOOL_ID } from "@/app/chat/tools/constants";
-import { updateTemperatureOverrideForChatSession } from "@/app/chat/lib";
 
 const CREDENTIAL_URL = "/api/manage/admin/credential";
 
@@ -362,22 +360,14 @@ export interface LlmOverride {
 export interface LlmOverrideManager {
   llmOverride: LlmOverride;
   updateLLMOverride: (newOverride: LlmOverride) => void;
-  temperature: number;
-  updateTemperature: (temperature: number) => void;
+  temperature: number | null;
+  updateTemperature: (temperature: number | null) => void;
   updateModelOverrideForChatSession: (chatSession?: ChatSession) => void;
   imageFilesPresent: boolean;
   updateImageFilesPresent: (present: boolean) => void;
   liveAssistant: Persona | null;
-  maxTemperature: number;
 }
 
-// Things to test
-// 1. User override
-// 2. User preference (defaults to system wide default if no preference set)
-// 3. Current assistant
-// 4. Current chat session
-// 5. Live assistant
-
 /*
 LLM Override is as follows (i.e. this order)
 - User override (explicitly set in the chat input bar)
@@ -396,20 +386,6 @@ Changes take place as
 - (uploadLLMOverride) User explicitly setting a model override (and we explicitly override and set the userSpecifiedOverride which we'll use in place of the user preferences unless overridden by an assistant)
 
 If we have a live assistant, we should use that model override
-
-Relevant test: `llm_ordering.spec.ts`.
-
-Temperature override is set as follows:
-- For existing chat sessions:
-  - If the user has previously overridden the temperature for a specific chat session, 
-    that value is persisted and used when the user returns to that chat.
-  - This persistence applies even if the temperature was set before sending the first message in the chat.
-- For new chat sessions:
-  - If the search tool is available, the default temperature is set to 0.
-  - If the search tool is not available, the default temperature is set to 0.5.
-
-This approach ensures that user preferences are maintained for existing chats while 
-providing appropriate defaults for new conversations based on the available tools.
 */
 
 export function useLlmOverride(
@@ -422,6 +398,11 @@ export function useLlmOverride(
   const [chatSession, setChatSession] = useState<ChatSession | null>(null);
 
   const llmOverrideUpdate = () => {
+    if (!chatSession && currentChatSession) {
+      setChatSession(currentChatSession || null);
+      return;
+    }
+
     if (liveAssistant?.llm_model_version_override) {
       setLlmOverride(
         getValidLlmOverride(liveAssistant.llm_model_version_override)
@@ -509,68 +490,24 @@ export function useLlmOverride(
     }
   };
 
-  const [temperature, setTemperature] = useState<number>(() => {
+  const [temperature, setTemperature] = useState<number | null>(0);
+
+  useEffect(() => {
     llmOverrideUpdate();
-
-    if (currentChatSession?.current_temperature_override != null) {
-      return Math.min(
-        currentChatSession.current_temperature_override,
-        isAnthropic(llmOverride.provider, llmOverride.modelName) ? 1.0 : 2.0
-      );
-    } else if (
-      liveAssistant?.tools.some((tool) => tool.name === SEARCH_TOOL_ID)
-    ) {
-      return 0;
-    }
-    return 0.5;
-  });
-
-  const maxTemperature = useMemo(() => {
-    return isAnthropic(llmOverride.provider, llmOverride.modelName) ? 1.0 : 2.0;
-  }, [llmOverride]);
-
-  useEffect(() => {
-    if (isAnthropic(llmOverride.provider, llmOverride.modelName)) {
-      const newTemperature = Math.min(temperature, 1.0);
-      setTemperature(newTemperature);
-      if (chatSession?.id) {
-        updateTemperatureOverrideForChatSession(chatSession.id, newTemperature);
-      }
-    }
-  }, [llmOverride]);
-
-  useEffect(() => {
-    if (!chatSession && currentChatSession) {
-      setChatSession(currentChatSession || null);
-      if (temperature) {
-        updateTemperatureOverrideForChatSession(
-          currentChatSession.id,
-          temperature
-        );
-      }
-      return;
-    }
-
-    if (currentChatSession?.current_temperature_override) {
-      setTemperature(currentChatSession.current_temperature_override);
-    } else if (
-      liveAssistant?.tools.some((tool) => tool.name === SEARCH_TOOL_ID)
-    ) {
-      setTemperature(0);
-    } else {
-      setTemperature(0.5);
-    }
   }, [liveAssistant, currentChatSession]);
 
-  const updateTemperature = (temperature: number) => {
+  useEffect(() => {
     if (isAnthropic(llmOverride.provider, llmOverride.modelName)) {
-      setTemperature((prevTemp) => Math.min(temperature, 1.0));
+      setTemperature((prevTemp) => Math.min(prevTemp ?? 0, 1.0));
+    }
+  }, [llmOverride]);
+
+  const updateTemperature = (temperature: number | null) => {
+    if (isAnthropic(llmOverride.provider, llmOverride.modelName)) {
+      setTemperature((prevTemp) => Math.min(temperature ?? 0, 1.0));
     } else {
       setTemperature(temperature);
     }
-    if (chatSession) {
-      updateTemperatureOverrideForChatSession(chatSession.id, temperature);
-    }
   };
 
   return {
@@ -582,7 +519,6 @@ export function useLlmOverride(
     imageFilesPresent,
     updateImageFilesPresent,
     liveAssistant: liveAssistant ?? null,
-    maxTemperature,
   };
 }
 

web/src/lib/types.ts

@@ -12,7 +12,6 @@ interface UserPreferences {
   recent_assistants: number[];
   auto_scroll: boolean | null;
   shortcut_enabled: boolean;
-  temperature_override_enabled: boolean;
 }
 
 export enum UserRole {

web/tests/e2e/admin_auth.setup.ts

@@ -1,9 +1,24 @@
 // dependency for all admin user tests
-import { test as setup } from "@playwright/test";
 
-setup("authenticate as admin", async ({ browser }) => {
-  const context = await browser.newContext({ storageState: "admin_auth.json" });
-  const page = await context.newPage();
+import { test as setup, expect } from "@playwright/test";
+import { TEST_CREDENTIALS } from "./constants";
+
+setup("authenticate", async ({ page }) => {
+  const { email, password } = TEST_CREDENTIALS;
+
   await page.goto("http://localhost:3000/chat");
+
+  await page.waitForURL("http://localhost:3000/auth/login?next=%2Fchat");
+
+  await expect(page).toHaveTitle("Onyx");
+
+  await page.fill("#email", email);
+  await page.fill("#password", password);
+
+  // Click the login button
+  await page.click('button[type="submit"]');
+
   await page.waitForURL("http://localhost:3000/chat");
+
+  await page.context().storageState({ path: "admin_auth.json" });
 });

web/tests/e2e/admin_oauth_redirect_uri.spec.ts

@@ -1,43 +1,65 @@
-import { test, expect } from "@playwright/test";
+import { test, expect } from "@chromatic-com/playwright";
 
-test.use({ storageState: "admin_auth.json" });
+test(
+  "Admin - OAuth Redirect - Missing Code",
+  {
+    tag: "@admin",
+  },
+  async ({ page }, testInfo) => {
+    await page.goto(
+      "http://localhost:3000/admin/connectors/slack/oauth/callback?state=xyz"
+    );
 
-test("Admin - OAuth Redirect - Missing Code", async ({ page }) => {
-  await page.goto(
-    "http://localhost:3000/admin/connectors/slack/oauth/callback?state=xyz"
-  );
+    await expect(page.locator("p.text-text-500")).toHaveText(
+      "Missing authorization code."
+    );
+  }
+);
 
-  await expect(page.locator("p.text-text-500")).toHaveText(
-    "Missing authorization code."
-  );
-});
+test(
+  "Admin - OAuth Redirect - Missing State",
+  {
+    tag: "@admin",
+  },
+  async ({ page }, testInfo) => {
+    await page.goto(
+      "http://localhost:3000/admin/connectors/slack/oauth/callback?code=123"
+    );
 
-test("Admin - OAuth Redirect - Missing State", async ({ page }) => {
-  await page.goto(
-    "http://localhost:3000/admin/connectors/slack/oauth/callback?code=123"
-  );
+    await expect(page.locator("p.text-text-500")).toHaveText(
+      "Missing state parameter."
+    );
+  }
+);
 
-  await expect(page.locator("p.text-text-500")).toHaveText(
-    "Missing state parameter."
-  );
-});
+test(
+  "Admin - OAuth Redirect - Invalid Connector",
+  {
+    tag: "@admin",
+  },
+  async ({ page }, testInfo) => {
+    await page.goto(
+      "http://localhost:3000/admin/connectors/invalid-connector/oauth/callback?code=123&state=xyz"
+    );
 
-test("Admin - OAuth Redirect - Invalid Connector", async ({ page }) => {
-  await page.goto(
-    "http://localhost:3000/admin/connectors/invalid-connector/oauth/callback?code=123&state=xyz"
-  );
+    await expect(page.locator("p.text-text-500")).toHaveText(
+      "invalid_connector is not a valid source type."
+    );
+  }
+);
 
-  await expect(page.locator("p.text-text-500")).toHaveText(
-    "invalid_connector is not a valid source type."
-  );
-});
+test(
+  "Admin - OAuth Redirect - No Session",
+  {
+    tag: "@admin",
+  },
+  async ({ page }, testInfo) => {
+    await page.goto(
+      "http://localhost:3000/admin/connectors/slack/oauth/callback?code=123&state=xyz"
+    );
 
-test("Admin - OAuth Redirect - No Session", async ({ page }) => {
-  await page.goto(
-    "http://localhost:3000/admin/connectors/slack/oauth/callback?code=123&state=xyz"
-  );
-
-  await expect(page.locator("p.text-text-500")).toHaveText(
-    "An error occurred during the OAuth process. Please try again."
-  );
-});
+    await expect(page.locator("p.text-text-500")).toHaveText(
+      "An error occurred during the OAuth process. Please try again."
+    );
+  }
+);