Merge branch 'main' into fix-daily-ux

When Cal.com events appear in an aggregated ICS feed, the same event
gets two different UIDs (one from Cal.com, one from Google Calendar).
This caused duplicate meetings to be created for the same time slot.

Add time-window dedup check in create_upcoming_meetings_for_event:
after verifying no meeting exists for the calendar_event_id, also check
if a meeting already exists for the same room + start_date + end_date.

.gitleaksignore

@@ -4,3 +4,4 @@ docs/docs/installation/daily-setup.md:curl-auth-header:277
 gpu/self_hosted/DEV_SETUP.md:curl-auth-header:74
 gpu/self_hosted/DEV_SETUP.md:curl-auth-header:83
 server/reflector/worker/process.py:generic-api-key:465
+server/reflector/worker/process.py:generic-api-key:594

gpu/self_hosted/Dockerfile

@@ -4,27 +4,31 @@ ENV PYTHONUNBUFFERED=1 \
     UV_LINK_MODE=copy \
     UV_NO_CACHE=1
 
+# patch until nvidia updates the sha1 repo
+ADD sequoia.config /etc/crypto-policies/back-ends/sequoia.config
+
 WORKDIR /tmp
-RUN apt-get update \
+RUN --mount=type=cache,target=/var/cache/apt,sharing=locked \
+ --mount=type=cache,target=/var/lib/apt,sharing=locked \
+ apt-get update \
  && apt-get install -y \
     ffmpeg \
     curl \
     ca-certificates \
     gnupg \
-    wget \
+    wget
- && apt-get clean
 # Add NVIDIA CUDA repo for Debian 12 (bookworm) and install cuDNN 9 for CUDA 12
 ADD https://developer.download.nvidia.com/compute/cuda/repos/debian12/x86_64/cuda-keyring_1.1-1_all.deb /cuda-keyring.deb
-RUN dpkg -i /cuda-keyring.deb \
+RUN --mount=type=cache,target=/var/cache/apt,sharing=locked \
+ --mount=type=cache,target=/var/lib/apt,sharing=locked \
+ dpkg -i /cuda-keyring.deb \
  && rm /cuda-keyring.deb \
  && apt-get update \
  && apt-get install -y --no-install-recommends \
     cuda-cudart-12-6 \
     libcublas-12-6 \
     libcudnn9-cuda-12 \
-    libcudnn9-dev-cuda-12 \
+    libcudnn9-dev-cuda-12
- && apt-get clean \
- && rm -rf /var/lib/apt/lists/*
 ADD https://astral.sh/uv/install.sh /uv-installer.sh
 RUN sh /uv-installer.sh && rm /uv-installer.sh
 ENV PATH="/root/.local/bin/:$PATH"
@@ -39,6 +43,13 @@ COPY ./app /app/app
 COPY ./main.py /app/
 COPY ./runserver.sh /app/
 
+# prevent uv failing with too many open files on big cpus
+ENV UV_CONCURRENT_INSTALLS=16
+
+# first install
+RUN --mount=type=cache,target=/root/.cache/uv \
+    uv sync --compile-bytecode --locked
+
 EXPOSE 8000
 
 CMD ["sh", "/app/runserver.sh"]

gpu/self_hosted/sequoia.config

@@ -0,0 +1,2 @@
+[hash_algorithms]
+sha1 = "always"

server/migrations/versions/3aa20b96d963_drop_use_celery_column.py

@@ -0,0 +1,35 @@
+"""drop_use_celery_column
+
+Revision ID: 3aa20b96d963
+Revises: e69f08ead8ea
+Create Date: 2026-02-05 10:12:44.065279
+
+"""
+
+from typing import Sequence, Union
+
+import sqlalchemy as sa
+from alembic import op
+
+# revision identifiers, used by Alembic.
+revision: str = "3aa20b96d963"
+down_revision: Union[str, None] = "e69f08ead8ea"
+branch_labels: Union[str, Sequence[str], None] = None
+depends_on: Union[str, Sequence[str], None] = None
+
+
+def upgrade() -> None:
+    with op.batch_alter_table("room", schema=None) as batch_op:
+        batch_op.drop_column("use_celery")
+
+
+def downgrade() -> None:
+    with op.batch_alter_table("room", schema=None) as batch_op:
+        batch_op.add_column(
+            sa.Column(
+                "use_celery",
+                sa.Boolean(),
+                server_default=sa.text("false"),
+                nullable=False,
+            )
+        )

server/reflector/db/meetings.py

@@ -346,6 +346,27 @@ class MeetingController:
             return None
         return Meeting(**result)
 
+    async def get_by_room_and_time_window(
+        self, room: Room, start_date: datetime, end_date: datetime
+    ) -> Meeting | None:
+        """Check if a meeting already exists for this room with the same time window."""
+        query = (
+            meetings.select()
+            .where(
+                sa.and_(
+                    meetings.c.room_id == room.id,
+                    meetings.c.start_date == start_date,
+                    meetings.c.end_date == end_date,
+                    meetings.c.is_active,
+                )
+            )
+            .limit(1)
+        )
+        result = await get_database().fetch_one(query)
+        if not result:
+            return None
+        return Meeting(**result)
+
     async def update_meeting(self, meeting_id: str, **kwargs):
         query = meetings.update().where(meetings.c.id == meeting_id).values(**kwargs)
         await get_database().execute(query)

server/reflector/db/rooms.py

@@ -57,12 +57,6 @@ rooms = sqlalchemy.Table(
         sqlalchemy.String,
         nullable=False,
     ),
-    sqlalchemy.Column(
-        "use_celery",
-        sqlalchemy.Boolean,
-        nullable=False,
-        server_default=false(),
-    ),
     sqlalchemy.Column(
         "skip_consent",
         sqlalchemy.Boolean,
@@ -97,7 +91,6 @@ class Room(BaseModel):
     ics_last_sync: datetime | None = None
     ics_last_etag: str | None = None
     platform: Platform = Field(default_factory=lambda: settings.DEFAULT_VIDEO_PLATFORM)
-    use_celery: bool = False
     skip_consent: bool = False
 
 

server/reflector/services/transcript_process.py

@@ -15,14 +15,10 @@ from hatchet_sdk.clients.rest.exceptions import ApiException, NotFoundException
 from hatchet_sdk.clients.rest.models import V1TaskStatus
 
 from reflector.db.recordings import recordings_controller
-from reflector.db.rooms import rooms_controller
 from reflector.db.transcripts import Transcript, transcripts_controller
 from reflector.hatchet.client import HatchetClientManager
 from reflector.logger import logger
 from reflector.pipelines.main_file_pipeline import task_pipeline_file_process
-from reflector.pipelines.main_multitrack_pipeline import (
-    task_pipeline_multitrack_process,
-)
 from reflector.utils.string import NonEmptyString
 
 
@@ -181,124 +177,98 @@ async def dispatch_transcript_processing(
     Returns AsyncResult for Celery tasks, None for Hatchet workflows.
     """
     if isinstance(config, MultitrackProcessingConfig):
-        use_celery = False
+        # Multitrack processing always uses Hatchet (no Celery fallback)
-        if config.room_id:
+        # First check if we can replay (outside transaction since it's read-only)
-            room = await rooms_controller.get_by_id(config.room_id)
+        transcript = await transcripts_controller.get_by_id(config.transcript_id)
-            use_celery = room.use_celery if room else False
+        if transcript and transcript.workflow_run_id and not force:
-
+            can_replay = await HatchetClientManager.can_replay(
-        use_hatchet = not use_celery
+                transcript.workflow_run_id
-
-        if use_celery:
-            logger.info(
-                "Room uses legacy Celery processing",
-                room_id=config.room_id,
-                transcript_id=config.transcript_id,
             )
-
+            if can_replay:
-        if use_hatchet:
+                await HatchetClientManager.replay_workflow(transcript.workflow_run_id)
-            # First check if we can replay (outside transaction since it's read-only)
+                logger.info(
-            transcript = await transcripts_controller.get_by_id(config.transcript_id)
+                    "Replaying Hatchet workflow",
-            if transcript and transcript.workflow_run_id and not force:
+                    workflow_id=transcript.workflow_run_id,
-                can_replay = await HatchetClientManager.can_replay(
-                    transcript.workflow_run_id
                 )
-                if can_replay:
+                return None
-                    await HatchetClientManager.replay_workflow(
+            else:
-                        transcript.workflow_run_id
+                # Workflow can't replay (CANCELLED, COMPLETED, or 404 deleted)
-                    )
+                # Log and proceed to start new workflow
-                    logger.info(
-                        "Replaying Hatchet workflow",
-                        workflow_id=transcript.workflow_run_id,
-                    )
-                    return None
-                else:
-                    # Workflow can't replay (CANCELLED, COMPLETED, or 404 deleted)
-                    # Log and proceed to start new workflow
-                    try:
-                        status = await HatchetClientManager.get_workflow_run_status(
-                            transcript.workflow_run_id
-                        )
-                        logger.info(
-                            "Old workflow not replayable, starting new",
-                            old_workflow_id=transcript.workflow_run_id,
-                            old_status=status.value,
-                        )
-                    except NotFoundException:
-                        # Workflow deleted from Hatchet but ID still in DB
-                        logger.info(
-                            "Old workflow not found in Hatchet, starting new",
-                            old_workflow_id=transcript.workflow_run_id,
-                        )
-
-            # Force: cancel old workflow if exists
-            if force and transcript and transcript.workflow_run_id:
-                try:
-                    await HatchetClientManager.cancel_workflow(
-                        transcript.workflow_run_id
-                    )
-                    logger.info(
-                        "Cancelled old workflow (--force)",
-                        workflow_id=transcript.workflow_run_id,
-                    )
-                except NotFoundException:
-                    logger.info(
-                        "Old workflow already deleted (--force)",
-                        workflow_id=transcript.workflow_run_id,
-                    )
-                await transcripts_controller.update(
-                    transcript, {"workflow_run_id": None}
-                )
-
-            # Re-fetch and check for concurrent dispatch (optimistic approach).
-            # No database lock - worst case is duplicate dispatch, but Hatchet
-            # workflows are idempotent so this is acceptable.
-            transcript = await transcripts_controller.get_by_id(config.transcript_id)
-            if transcript and transcript.workflow_run_id:
-                # Another process started a workflow between validation and now
                 try:
                     status = await HatchetClientManager.get_workflow_run_status(
                         transcript.workflow_run_id
                     )
-                    if status in (V1TaskStatus.RUNNING, V1TaskStatus.QUEUED):
+                    logger.info(
-                        logger.info(
+                        "Old workflow not replayable, starting new",
-                            "Concurrent workflow detected, skipping dispatch",
+                        old_workflow_id=transcript.workflow_run_id,
-                            workflow_id=transcript.workflow_run_id,
+                        old_status=status.value,
-                        )
+                    )
-                        return None
+                except NotFoundException:
-                except ApiException:
+                    # Workflow deleted from Hatchet but ID still in DB
-                    # Workflow might be gone (404) or API issue - proceed with new workflow
+                    logger.info(
-                    pass
+                        "Old workflow not found in Hatchet, starting new",
+                        old_workflow_id=transcript.workflow_run_id,
+                    )
 
-            workflow_id = await HatchetClientManager.start_workflow(
+        # Force: cancel old workflow if exists
-                workflow_name="DiarizationPipeline",
+        if force and transcript and transcript.workflow_run_id:
-                input_data={
+            try:
-                    "recording_id": config.recording_id,
+                await HatchetClientManager.cancel_workflow(transcript.workflow_run_id)
-                    "tracks": [{"s3_key": k} for k in config.track_keys],
+                logger.info(
-                    "bucket_name": config.bucket_name,
+                    "Cancelled old workflow (--force)",
-                    "transcript_id": config.transcript_id,
+                    workflow_id=transcript.workflow_run_id,
-                    "room_id": config.room_id,
+                )
-                },
+            except NotFoundException:
-                additional_metadata={
+                logger.info(
-                    "transcript_id": config.transcript_id,
+                    "Old workflow already deleted (--force)",
-                    "recording_id": config.recording_id,
+                    workflow_id=transcript.workflow_run_id,
-                    "daily_recording_id": config.recording_id,
+                )
-                },
+            await transcripts_controller.update(transcript, {"workflow_run_id": None})
+
+        # Re-fetch and check for concurrent dispatch (optimistic approach).
+        # No database lock - worst case is duplicate dispatch, but Hatchet
+        # workflows are idempotent so this is acceptable.
+        transcript = await transcripts_controller.get_by_id(config.transcript_id)
+        if transcript and transcript.workflow_run_id:
+            # Another process started a workflow between validation and now
+            try:
+                status = await HatchetClientManager.get_workflow_run_status(
+                    transcript.workflow_run_id
+                )
+                if status in (V1TaskStatus.RUNNING, V1TaskStatus.QUEUED):
+                    logger.info(
+                        "Concurrent workflow detected, skipping dispatch",
+                        workflow_id=transcript.workflow_run_id,
+                    )
+                    return None
+            except ApiException:
+                # Workflow might be gone (404) or API issue - proceed with new workflow
+                pass
+
+        workflow_id = await HatchetClientManager.start_workflow(
+            workflow_name="DiarizationPipeline",
+            input_data={
+                "recording_id": config.recording_id,
+                "tracks": [{"s3_key": k} for k in config.track_keys],
+                "bucket_name": config.bucket_name,
+                "transcript_id": config.transcript_id,
+                "room_id": config.room_id,
+            },
+            additional_metadata={
+                "transcript_id": config.transcript_id,
+                "recording_id": config.recording_id,
+                "daily_recording_id": config.recording_id,
+            },
+        )
+
+        if transcript:
+            await transcripts_controller.update(
+                transcript, {"workflow_run_id": workflow_id}
             )
 
-            if transcript:
+        logger.info("Hatchet workflow dispatched", workflow_id=workflow_id)
-                await transcripts_controller.update(
+        return None
-                    transcript, {"workflow_run_id": workflow_id}
-                )
 
-            logger.info("Hatchet workflow dispatched", workflow_id=workflow_id)
-            return None
-
-        # Celery pipeline (durable workflows disabled)
-        return task_pipeline_multitrack_process.delay(
-            transcript_id=config.transcript_id,
-            bucket_name=config.bucket_name,
-            track_keys=config.track_keys,
-        )
     elif isinstance(config, FileProcessingConfig):
         return task_pipeline_file_process.delay(transcript_id=config.transcript_id)
     else:

server/reflector/settings.py

@@ -1,7 +1,7 @@
 from pydantic.types import PositiveInt
 from pydantic_settings import BaseSettings, SettingsConfigDict
 
-from reflector.schemas.platform import WHEREBY_PLATFORM, Platform
+from reflector.schemas.platform import DAILY_PLATFORM, Platform
 from reflector.utils.string import NonEmptyString
 
 
@@ -155,7 +155,7 @@ class Settings(BaseSettings):
         None  # Webhook UUID for this environment. Not used by production code
     )
     # Platform Configuration
-    DEFAULT_VIDEO_PLATFORM: Platform = WHEREBY_PLATFORM
+    DEFAULT_VIDEO_PLATFORM: Platform = DAILY_PLATFORM
 
     # Zulip integration
     ZULIP_REALM: str | None = None

server/reflector/worker/ics_sync.py

@@ -5,7 +5,7 @@ from celery import shared_task
 from celery.utils.log import get_task_logger
 
 from reflector.asynctask import asynctask
-from reflector.db.calendar_events import calendar_events_controller
+from reflector.db.calendar_events import CalendarEvent, calendar_events_controller
 from reflector.db.meetings import meetings_controller
 from reflector.db.rooms import Room, rooms_controller
 from reflector.redis_cache import RedisAsyncLock
@@ -83,10 +83,9 @@ def _should_sync(room) -> bool:
     return time_since_sync.total_seconds() >= room.ics_fetch_interval
 
 
-MEETING_DEFAULT_DURATION = timedelta(hours=1)
+async def create_upcoming_meetings_for_event(
-
+    event: CalendarEvent, create_window: datetime, room: Room
-
+):
-async def create_upcoming_meetings_for_event(event, create_window, room: Room):
     if event.start_time <= create_window:
         return
     existing_meeting = await meetings_controller.get_by_calendar_event(event.id, room)
@@ -94,6 +93,21 @@ async def create_upcoming_meetings_for_event(event, create_window, room: Room):
     if existing_meeting:
         return
 
+    # Prevent duplicate meetings from aggregated calendar feeds
+    # (e.g. same event appears with different UIDs from Cal.com and Google Calendar)
+    end_date = event.end_time
+    existing_by_time = await meetings_controller.get_by_room_and_time_window(
+        room, event.start_time, end_date
+    )
+    if existing_by_time:
+        logger.info(
+            "Skipping duplicate calendar event - meeting already exists for this time window",
+            room_id=room.id,
+            event_id=event.id,
+            existing_meeting_id=existing_by_time.id,
+        )
+        return
+
     logger.info(
         "Pre-creating meeting for calendar event",
         room_id=room.id,
@@ -102,8 +116,6 @@ async def create_upcoming_meetings_for_event(event, create_window, room: Room):
     )
 
     try:
-        end_date = event.end_time or (event.start_time + MEETING_DEFAULT_DURATION)
-
         client = create_platform_client(room.platform)
 
         meeting_data = await client.create_meeting(

server/reflector/worker/process.py

@@ -27,9 +27,6 @@ from reflector.db.transcripts import (
 from reflector.hatchet.client import HatchetClientManager
 from reflector.pipelines.main_file_pipeline import task_pipeline_file_process
 from reflector.pipelines.main_live_pipeline import asynctask
-from reflector.pipelines.main_multitrack_pipeline import (
-    task_pipeline_multitrack_process,
-)
 from reflector.pipelines.topic_processing import EmptyPipeline
 from reflector.processors import AudioFileWriterProcessor
 from reflector.processors.audio_waveform_processor import AudioWaveformProcessor
@@ -351,49 +348,29 @@ async def _process_multitrack_recording_inner(
             room_id=room.id,
         )
 
-    use_celery = room and room.use_celery
+    # Multitrack processing always uses Hatchet (no Celery fallback)
-    use_hatchet = not use_celery
+    workflow_id = await HatchetClientManager.start_workflow(
-
+        workflow_name="DiarizationPipeline",
-    if use_celery:
+        input_data={
-        logger.info(
+            "recording_id": recording_id,
-            "Room uses legacy Celery processing",
+            "tracks": [{"s3_key": k} for k in filter_cam_audio_tracks(track_keys)],
-            room_id=room.id,
+            "bucket_name": bucket_name,
-            transcript_id=transcript.id,
+            "transcript_id": transcript.id,
-        )
+            "room_id": room.id,
-
+        },
-    if use_hatchet:
+        additional_metadata={
-        workflow_id = await HatchetClientManager.start_workflow(
+            "transcript_id": transcript.id,
-            workflow_name="DiarizationPipeline",
+            "recording_id": recording_id,
-            input_data={
+            "daily_recording_id": recording_id,
-                "recording_id": recording_id,
+        },
-                "tracks": [{"s3_key": k} for k in filter_cam_audio_tracks(track_keys)],
-                "bucket_name": bucket_name,
-                "transcript_id": transcript.id,
-                "room_id": room.id,
-            },
-            additional_metadata={
-                "transcript_id": transcript.id,
-                "recording_id": recording_id,
-                "daily_recording_id": recording_id,
-            },
-        )
-        logger.info(
-            "Started Hatchet workflow",
-            workflow_id=workflow_id,
-            transcript_id=transcript.id,
-        )
-
-        await transcripts_controller.update(
-            transcript, {"workflow_run_id": workflow_id}
-        )
-        return
-
-    # Celery pipeline (runs when durable workflows disabled)
-    task_pipeline_multitrack_process.delay(
-        transcript_id=transcript.id,
-        bucket_name=bucket_name,
-        track_keys=filter_cam_audio_tracks(track_keys),
     )
+    logger.info(
+        "Started Hatchet workflow",
+        workflow_id=workflow_id,
+        transcript_id=transcript.id,
+    )
+
+    await transcripts_controller.update(transcript, {"workflow_run_id": workflow_id})
 
 
 @shared_task
@@ -1072,66 +1049,43 @@ async def reprocess_failed_daily_recordings():
                 )
                 continue
 
-            use_celery = room and room.use_celery
+            # Multitrack reprocessing always uses Hatchet (no Celery fallback)
-            use_hatchet = not use_celery
+            if not transcript:
-
+                logger.warning(
-            if use_hatchet:
+                    "No transcript for Hatchet reprocessing, skipping",
-                if not transcript:
-                    logger.warning(
-                        "No transcript for Hatchet reprocessing, skipping",
-                        recording_id=recording.id,
-                    )
-                    continue
-
-                workflow_id = await HatchetClientManager.start_workflow(
-                    workflow_name="DiarizationPipeline",
-                    input_data={
-                        "recording_id": recording.id,
-                        "tracks": [
-                            {"s3_key": k}
-                            for k in filter_cam_audio_tracks(recording.track_keys)
-                        ],
-                        "bucket_name": bucket_name,
-                        "transcript_id": transcript.id,
-                        "room_id": room.id if room else None,
-                    },
-                    additional_metadata={
-                        "transcript_id": transcript.id,
-                        "recording_id": recording.id,
-                        "reprocess": True,
-                    },
-                )
-                await transcripts_controller.update(
-                    transcript, {"workflow_run_id": workflow_id}
-                )
-
-                logger.info(
-                    "Queued Daily recording for Hatchet reprocessing",
                     recording_id=recording.id,
-                    workflow_id=workflow_id,
-                    room_name=meeting.room_name,
-                    track_count=len(recording.track_keys),
-                )
-            else:
-                logger.info(
-                    "Queueing Daily recording for Celery reprocessing",
-                    recording_id=recording.id,
-                    room_name=meeting.room_name,
-                    track_count=len(recording.track_keys),
-                    transcript_status=transcript.status if transcript else None,
                 )
+                continue
 
-                # For reprocessing, pass actual recording time (though it's ignored - see _process_multitrack_recording_inner)
+            workflow_id = await HatchetClientManager.start_workflow(
-                # Reprocessing uses recording.meeting_id directly instead of time-based matching
+                workflow_name="DiarizationPipeline",
-                recording_start_ts = int(recording.recorded_at.timestamp())
+                input_data={
+                    "recording_id": recording.id,
+                    "tracks": [
+                        {"s3_key": k}
+                        for k in filter_cam_audio_tracks(recording.track_keys)
+                    ],
+                    "bucket_name": bucket_name,
+                    "transcript_id": transcript.id,
+                    "room_id": room.id if room else None,
+                },
+                additional_metadata={
+                    "transcript_id": transcript.id,
+                    "recording_id": recording.id,
+                    "reprocess": True,
+                },
+            )
+            await transcripts_controller.update(
+                transcript, {"workflow_run_id": workflow_id}
+            )
 
-                process_multitrack_recording.delay(
+            logger.info(
-                    bucket_name=bucket_name,
+                "Queued Daily recording for Hatchet reprocessing",
-                    daily_room_name=meeting.room_name,
+                recording_id=recording.id,
-                    recording_id=recording.id,
+                workflow_id=workflow_id,
-                    track_keys=recording.track_keys,
+                room_name=meeting.room_name,
-                    recording_start_ts=recording_start_ts,
+                track_count=len(recording.track_keys),
-                )
+            )
 
             reprocessed_count += 1
 

server/reflector/ws_manager.py

@@ -11,7 +11,6 @@ broadcast messages to all connected websockets.
 
 import asyncio
 import json
-import threading
 
 import redis.asyncio as redis
 from fastapi import WebSocket
@@ -98,6 +97,7 @@ class WebsocketManager:
 
     async def _pubsub_data_reader(self, pubsub_subscriber):
         while True:
+            # timeout=1.0 prevents tight CPU loop when no messages available
             message = await pubsub_subscriber.get_message(
                 ignore_subscribe_messages=True
             )
@@ -109,29 +109,38 @@ class WebsocketManager:
                     await socket.send_json(data)
 
 
+# Process-global singleton to ensure only one WebsocketManager instance exists.
+# Multiple instances would cause resource leaks and CPU issues.
+_ws_manager: WebsocketManager | None = None
+
+
 def get_ws_manager() -> WebsocketManager:
     """
-    Returns the WebsocketManager instance for managing websockets.
+    Returns the global WebsocketManager singleton.
 
-    This function initializes and returns the WebsocketManager instance,
+    Creates instance on first call, subsequent calls return cached instance.
-    which is responsible for managing websockets and handling websocket
+    Thread-safe via GIL. Concurrent initialization may create duplicate
-    connections.
+    instances but last write wins (acceptable for this use case).
 
     Returns:
-        WebsocketManager: The initialized WebsocketManager instance.
+        WebsocketManager: The global WebsocketManager instance.
-
-    Raises:
-        ImportError: If the 'reflector.settings' module cannot be imported.
-        RedisConnectionError: If there is an error connecting to the Redis server.
     """
-    local = threading.local()
+    global _ws_manager
-    if hasattr(local, "ws_manager"):
-        return local.ws_manager
 
+    if _ws_manager is not None:
+        return _ws_manager
+
+    # No lock needed - GIL makes this safe enough
+    # Worst case: race creates two instances, last assignment wins
     pubsub_client = RedisPubSubManager(
         host=settings.REDIS_HOST,
         port=settings.REDIS_PORT,
     )
-    ws_manager = WebsocketManager(pubsub_client=pubsub_client)
+    _ws_manager = WebsocketManager(pubsub_client=pubsub_client)
-    local.ws_manager = ws_manager
+    return _ws_manager
-    return ws_manager
+
+
+def reset_ws_manager() -> None:
+    """Reset singleton for testing. DO NOT use in production."""
+    global _ws_manager
+    _ws_manager = None

server/tests/conftest.py

@@ -1,11 +1,10 @@
 import os
 from contextlib import asynccontextmanager
-from tempfile import NamedTemporaryFile
 from unittest.mock import patch
 
 import pytest
 
-from reflector.schemas.platform import WHEREBY_PLATFORM
+from reflector.schemas.platform import DAILY_PLATFORM, WHEREBY_PLATFORM
 
 
 @pytest.fixture(scope="session", autouse=True)
@@ -15,6 +14,7 @@ def register_mock_platform():
     from reflector.video_platforms.registry import register_platform
 
     register_platform(WHEREBY_PLATFORM, MockPlatformClient)
+    register_platform(DAILY_PLATFORM, MockPlatformClient)
     yield
 
 
@@ -333,11 +333,14 @@ def celery_enable_logging():
 
 @pytest.fixture(scope="session")
 def celery_config():
-    with NamedTemporaryFile() as f:
+    redis_host = os.environ.get("REDIS_HOST", "localhost")
-        yield {
+    redis_port = os.environ.get("REDIS_PORT", "6379")
-            "broker_url": "memory://",
+    # Use db 2 to avoid conflicts with main app
-            "result_backend": f"db+sqlite:///{f.name}",
+    redis_url = f"redis://{redis_host}:{redis_port}/2"
-        }
+    yield {
+        "broker_url": redis_url,
+        "result_backend": redis_url,
+    }
 
 
 @pytest.fixture(scope="session")
@@ -370,9 +373,12 @@ async def ws_manager_in_memory(monkeypatch):
         def __init__(self, queue: asyncio.Queue):
             self.queue = queue
 
-        async def get_message(self, ignore_subscribe_messages: bool = True):
+        async def get_message(
+            self, ignore_subscribe_messages: bool = True, timeout: float | None = None
+        ):
+            wait_timeout = timeout if timeout is not None else 0.05
             try:
-                return await asyncio.wait_for(self.queue.get(), timeout=0.05)
+                return await asyncio.wait_for(self.queue.get(), timeout=wait_timeout)
             except Exception:
                 return None
 

server/tests/test_ics_dedup.py

@@ -0,0 +1,190 @@
+from datetime import datetime, timedelta, timezone
+from unittest.mock import AsyncMock, patch
+
+import pytest
+
+from reflector.db import get_database
+from reflector.db.calendar_events import CalendarEvent, calendar_events_controller
+from reflector.db.meetings import meetings
+from reflector.db.rooms import rooms_controller
+from reflector.worker.ics_sync import create_upcoming_meetings_for_event
+
+
+@pytest.mark.asyncio
+async def test_duplicate_calendar_event_does_not_create_duplicate_meeting():
+    """When an aggregated ICS feed contains the same event with different UIDs
+    (e.g. Cal.com UID + Google Calendar UUID), only one meeting should be created."""
+
+    room = await rooms_controller.add(
+        name="dedup-test-room",
+        user_id="test-user",
+        zulip_auto_post=False,
+        zulip_stream="",
+        zulip_topic="",
+        is_locked=False,
+        room_mode="normal",
+        recording_type="cloud",
+        recording_trigger="automatic-2nd-participant",
+        is_shared=False,
+        ics_url="https://calendar.example.com/dedup.ics",
+        ics_enabled=True,
+    )
+
+    now = datetime.now(timezone.utc)
+    start_time = now + timedelta(hours=1)
+    end_time = now + timedelta(hours=2)
+
+    # Create first calendar event (Cal.com UID)
+    event1 = await calendar_events_controller.upsert(
+        CalendarEvent(
+            room_id=room.id,
+            ics_uid="abc123@Cal.com",
+            title="Team Standup",
+            start_time=start_time,
+            end_time=end_time,
+        )
+    )
+
+    # create_window must be before start_time for the function to proceed
+    create_window = now - timedelta(minutes=6)
+
+    # Create meeting for event1
+    with patch(
+        "reflector.worker.ics_sync.create_platform_client"
+    ) as mock_platform_factory:
+        mock_client = AsyncMock()
+
+        async def mock_create_meeting_1(room_name_prefix, *, end_date, room):
+            return AsyncMock(
+                meeting_id="meeting-1",
+                room_name="dedup-test-room-abc",
+                room_url="https://mock.video/dedup-test-room-abc",
+                host_room_url="https://mock.video/dedup-test-room-abc?host=true",
+            )
+
+        mock_client.create_meeting = mock_create_meeting_1
+        mock_client.upload_logo = AsyncMock()
+        mock_platform_factory.return_value = mock_client
+
+        await create_upcoming_meetings_for_event(event1, create_window, room)
+
+    # Verify meeting was created
+    results = await get_database().fetch_all(
+        meetings.select().where(meetings.c.room_id == room.id)
+    )
+    assert len(results) == 1, f"Expected 1 meeting, got {len(results)}"
+
+    # Create second calendar event with different UID but same time window (Google Calendar UUID)
+    event2 = await calendar_events_controller.upsert(
+        CalendarEvent(
+            room_id=room.id,
+            ics_uid="550e8400-e29b-41d4-a716-446655440000",
+            title="Team Standup",
+            start_time=start_time,
+            end_time=end_time,
+        )
+    )
+
+    # Try to create meeting for event2 - should be skipped due to dedup
+    with patch(
+        "reflector.worker.ics_sync.create_platform_client"
+    ) as mock_platform_factory:
+        mock_client = AsyncMock()
+        create_meeting_called = False
+
+        async def mock_create_meeting_2(room_name_prefix, *, end_date, room):
+            nonlocal create_meeting_called
+            create_meeting_called = True
+
+        mock_client.create_meeting = mock_create_meeting_2
+        mock_client.upload_logo = AsyncMock()
+        mock_platform_factory.return_value = mock_client
+
+        await create_upcoming_meetings_for_event(event2, create_window, room)
+
+        # Platform client should NOT have been called for the duplicate
+        assert (
+            not create_meeting_called
+        ), "create_meeting should not be called for duplicate"
+
+    # Verify still only 1 meeting
+    results = await get_database().fetch_all(
+        meetings.select().where(meetings.c.room_id == room.id)
+    )
+    assert len(results) == 1, f"Expected 1 meeting after dedup, got {len(results)}"
+
+
+@pytest.mark.asyncio
+async def test_different_time_windows_create_separate_meetings():
+    """Events at different times should create separate meetings, even if titles match."""
+
+    room = await rooms_controller.add(
+        name="dedup-diff-time-room",
+        user_id="test-user",
+        zulip_auto_post=False,
+        zulip_stream="",
+        zulip_topic="",
+        is_locked=False,
+        room_mode="normal",
+        recording_type="cloud",
+        recording_trigger="automatic-2nd-participant",
+        is_shared=False,
+        ics_url="https://calendar.example.com/dedup2.ics",
+        ics_enabled=True,
+    )
+
+    now = datetime.now(timezone.utc)
+    create_window = now - timedelta(minutes=6)
+
+    # Event 1: 1-2pm
+    event1 = await calendar_events_controller.upsert(
+        CalendarEvent(
+            room_id=room.id,
+            ics_uid="event-morning@Cal.com",
+            title="Team Standup",
+            start_time=now + timedelta(hours=1),
+            end_time=now + timedelta(hours=2),
+        )
+    )
+
+    # Event 2: 3-4pm (different time)
+    event2 = await calendar_events_controller.upsert(
+        CalendarEvent(
+            room_id=room.id,
+            ics_uid="event-afternoon@Cal.com",
+            title="Team Standup",
+            start_time=now + timedelta(hours=3),
+            end_time=now + timedelta(hours=4),
+        )
+    )
+
+    with patch(
+        "reflector.worker.ics_sync.create_platform_client"
+    ) as mock_platform_factory:
+        mock_client = AsyncMock()
+
+        call_count = 0
+
+        async def mock_create_meeting(room_name_prefix, *, end_date, room):
+            nonlocal call_count
+            call_count += 1
+            return AsyncMock(
+                meeting_id=f"meeting-{call_count}",
+                room_name=f"dedup-diff-time-room-{call_count}",
+                room_url=f"https://mock.video/dedup-diff-time-room-{call_count}",
+                host_room_url=f"https://mock.video/dedup-diff-time-room-{call_count}?host=true",
+            )
+
+        mock_client.create_meeting = mock_create_meeting
+        mock_client.upload_logo = AsyncMock()
+        mock_platform_factory.return_value = mock_client
+
+        await create_upcoming_meetings_for_event(event1, create_window, room)
+        await create_upcoming_meetings_for_event(event2, create_window, room)
+
+    results = await get_database().fetch_all(
+        meetings.select().where(meetings.c.room_id == room.id)
+    )
+    assert (
+        len(results) == 2
+    ), f"Expected 2 meetings for different times, got {len(results)}"

server/tests/test_transcripts_process.py

@@ -1,6 +1,6 @@
 import asyncio
 import time
-from unittest.mock import patch
+from unittest.mock import AsyncMock, patch
 
 import pytest
 from httpx import ASGITransport, AsyncClient
@@ -142,17 +142,17 @@ async def test_whereby_recording_uses_file_pipeline(client):
             "reflector.services.transcript_process.task_pipeline_file_process"
         ) as mock_file_pipeline,
         patch(
-            "reflector.services.transcript_process.task_pipeline_multitrack_process"
+            "reflector.services.transcript_process.HatchetClientManager"
-        ) as mock_multitrack_pipeline,
+        ) as mock_hatchet,
     ):
         response = await client.post(f"/transcripts/{transcript.id}/process")
 
         assert response.status_code == 200
         assert response.json()["status"] == "ok"
 
-        # Whereby recordings should use file pipeline
+        # Whereby recordings should use file pipeline, not Hatchet
         mock_file_pipeline.delay.assert_called_once_with(transcript_id=transcript.id)
-        mock_multitrack_pipeline.delay.assert_not_called()
+        mock_hatchet.start_workflow.assert_not_called()
 
 
 @pytest.mark.usefixtures("setup_database")
@@ -177,8 +177,6 @@ async def test_dailyco_recording_uses_multitrack_pipeline(client):
         recording_trigger="automatic-2nd-participant",
         is_shared=False,
     )
-    # Force Celery backend for test
-    await rooms_controller.update(room, {"use_celery": True})
 
     transcript = await transcripts_controller.add(
         "",
@@ -213,18 +211,23 @@ async def test_dailyco_recording_uses_multitrack_pipeline(client):
             "reflector.services.transcript_process.task_pipeline_file_process"
         ) as mock_file_pipeline,
         patch(
-            "reflector.services.transcript_process.task_pipeline_multitrack_process"
+            "reflector.services.transcript_process.HatchetClientManager"
-        ) as mock_multitrack_pipeline,
+        ) as mock_hatchet,
     ):
+        mock_hatchet.start_workflow = AsyncMock(return_value="test-workflow-id")
+
         response = await client.post(f"/transcripts/{transcript.id}/process")
 
         assert response.status_code == 200
         assert response.json()["status"] == "ok"
 
-        # Daily.co multitrack recordings should use multitrack pipeline
+        # Daily.co multitrack recordings should use Hatchet workflow
-        mock_multitrack_pipeline.delay.assert_called_once_with(
+        mock_hatchet.start_workflow.assert_called_once()
-            transcript_id=transcript.id,
+        call_kwargs = mock_hatchet.start_workflow.call_args.kwargs
-            bucket_name="daily-bucket",
+        assert call_kwargs["workflow_name"] == "DiarizationPipeline"
-            track_keys=track_keys,
+        assert call_kwargs["input_data"]["transcript_id"] == transcript.id
-        )
+        assert call_kwargs["input_data"]["bucket_name"] == "daily-bucket"
+        assert call_kwargs["input_data"]["tracks"] == [
+            {"s3_key": k} for k in track_keys
+        ]
         mock_file_pipeline.delay.assert_not_called()

server/tests/test_transcripts_rtc_ws.py

@@ -115,9 +115,7 @@ def appserver(tmpdir, setup_database, celery_session_app, celery_session_worker)
     settings.DATA_DIR = DATA_DIR
 
 
-@pytest.fixture(scope="session")
+# Using celery_includes from conftest.py which includes both pipelines
-def celery_includes():
-    return ["reflector.pipelines.main_live_pipeline"]
 
 
 @pytest.mark.usefixtures("setup_database")

server/tests/test_user_websocket_auth.py

@@ -56,7 +56,12 @@ def appserver_ws_user(setup_database):
 
     if server_instance:
         server_instance.should_exit = True
-        server_thread.join(timeout=30)
+        server_thread.join(timeout=2.0)
+
+    # Reset global singleton for test isolation
+    from reflector.ws_manager import reset_ws_manager
+
+    reset_ws_manager()
 
 
 @pytest.fixture(autouse=True)
@@ -133,6 +138,8 @@ async def test_user_ws_accepts_valid_token_and_receives_events(appserver_ws_user
 
     # Connect and then trigger an event via HTTP create
     async with aconnect_ws(base_ws, subprotocols=subprotocols) as ws:
+        await asyncio.sleep(0.2)
+
         # Emit an event to the user's room via a standard HTTP action
         from httpx import AsyncClient
 
@@ -150,6 +157,7 @@ async def test_user_ws_accepts_valid_token_and_receives_events(appserver_ws_user
             "email": "user-abc@example.com",
         }
 
+        # Use in-memory client (global singleton makes it share ws_manager)
         async with AsyncClient(app=app, base_url=f"http://{host}:{port}/v1") as ac:
             # Create a transcript as this user so that the server publishes TRANSCRIPT_CREATED to user room
             resp = await ac.post("/transcripts", json={"name": "WS Test"})

www/app/[roomName]/components/DailyRoom.tsx

@@ -8,7 +8,7 @@ import {
   useRef,
   useState,
 } from "react";
-import { Box, Spinner, Center, Text } from "@chakra-ui/react";
+import { Box, Spinner, Center, Text, Button, VStack } from "@chakra-ui/react";
 import { useRouter, useParams } from "next/navigation";
 import DailyIframe, {
   DailyCall,
@@ -16,10 +16,13 @@ import DailyIframe, {
   DailyCustomTrayButton,
   DailyCustomTrayButtons,
   DailyEventObjectCustomButtonClick,
+  DailyEventObjectFatalError,
+  DailyFatalErrorType,
   DailyFactoryOptions,
   DailyParticipantsObject,
 } from "@daily-co/daily-js";
 import type { components } from "../../reflector-api";
+import { printApiError, ApiError } from "../../api/_error";
 import { useAuth } from "../../lib/AuthProvider";
 import { useConsentDialog } from "../../lib/consent";
 import {
@@ -45,6 +48,63 @@ const RAW_TRACKS_NAMESPACE = "a1b2c3d4-e5f6-7890-abcd-ef1234567890";
 const RECORDING_START_DELAY_MS = 2000;
 const RECORDING_START_MAX_RETRIES = 5;
 
+const FATAL_ERROR_MESSAGES: Partial<
+  Record<DailyFatalErrorType, { message: string; rejoinable?: boolean }>
+> = {
+  "connection-error": {
+    message: "Connection lost. Please check your network.",
+    rejoinable: true,
+  },
+  "exp-room": { message: "The meeting time has ended." },
+  "exp-token": { message: "Your session has expired.", rejoinable: true },
+  ejected: { message: "You were removed from this meeting." },
+  "meeting-full": { message: "This meeting is full." },
+  "not-allowed": { message: "You are not allowed to join this meeting." },
+  "nbf-room": { message: "This meeting hasn't started yet." },
+  "nbf-token": { message: "This meeting hasn't started yet." },
+  "no-room": { message: "This room does not exist." },
+  "end-of-life": { message: "This meeting room is no longer available." },
+};
+
+function FatalErrorScreen({
+  error,
+  roomName,
+}: {
+  error: FatalError;
+  roomName: string;
+}) {
+  const router = useRouter();
+  const info =
+    error.type !== "unknown" ? FATAL_ERROR_MESSAGES[error.type] : undefined;
+  const message = info?.message ?? `Something went wrong: ${error.message}`;
+  const rejoinable = info?.rejoinable ?? false;
+
+  return (
+    <Center width="100vw" height="100vh">
+      <VStack gap={4}>
+        <Text color="red.500">{message}</Text>
+        {rejoinable ? (
+          <>
+            <Button onClick={() => window.location.reload()}>
+              Try Rejoining
+            </Button>
+            <Button
+              variant="outline"
+              onClick={() => router.push(`/${roomName}`)}
+            >
+              Leave
+            </Button>
+          </>
+        ) : (
+          <Button onClick={() => router.push(`/${roomName}`)}>
+            Back to Room
+          </Button>
+        )}
+      </VStack>
+    </Center>
+  );
+}
+
 type Meeting = components["schemas"]["Meeting"];
 type Room = components["schemas"]["RoomDetails"];
 
@@ -82,6 +142,8 @@ const USE_FRAME_INIT_STATE = {
   joined: false as boolean,
 } as const;
 
+type FatalError = { type: DailyFatalErrorType | "unknown"; message: string };
+
 // Daily js and not Daily react used right now because daily-js allows for prebuild interface vs. -react is customizable but has no nice defaults
 const useFrame = (
   container: HTMLDivElement | null,
@@ -89,6 +151,7 @@ const useFrame = (
     onLeftMeeting: () => void;
     onCustomButtonClick: (ev: DailyEventObjectCustomButtonClick) => void;
     onJoinMeeting: () => void;
+    onError: (ev: DailyEventObjectFatalError) => void;
   },
 ) => {
   const [{ frame, joined }, setState] = useState(USE_FRAME_INIT_STATE);
@@ -134,6 +197,7 @@ const useFrame = (
     if (!frame) return;
     frame.on("left-meeting", cbs.onLeftMeeting);
     frame.on("custom-button-click", cbs.onCustomButtonClick);
+    frame.on("error", cbs.onError);
     const joinCb = () => {
       if (!frame) {
         console.error("frame is null in joined-meeting callback");
@@ -145,6 +209,7 @@ const useFrame = (
     return () => {
       frame.off("left-meeting", cbs.onLeftMeeting);
       frame.off("custom-button-click", cbs.onCustomButtonClick);
+      frame.off("error", cbs.onError);
       frame.off("joined-meeting", joinCb);
     };
   }, [frame, cbs]);
@@ -188,6 +253,7 @@ export default function DailyRoom({ meeting, room }: DailyRoomProps) {
   const joinMutation = useRoomJoinMeeting();
   const startRecordingMutation = useMeetingStartRecording();
   const [joinedMeeting, setJoinedMeeting] = useState<Meeting | null>(null);
+  const [fatalError, setFatalError] = useState<FatalError | null>(null);
 
   // Generate deterministic instanceIds so all participants use SAME IDs
   const cloudInstanceId = parseNonEmptyString(meeting.id);
@@ -234,8 +300,18 @@ export default function DailyRoom({ meeting, room }: DailyRoomProps) {
   const roomUrl = joinedMeeting?.room_url;
 
   const handleLeave = useCallback(() => {
+    // If a fatal error occurred, don't redirect — let the error UI show
+    if (fatalError) return;
     router.push("/browse");
-  }, [router]);
+  }, [router, fatalError]);
+
+  const handleError = useCallback((ev: DailyEventObjectFatalError) => {
+    const error: FatalError = {
+      type: ev.error?.type ?? "unknown",
+      message: ev.errorMsg,
+    };
+    setFatalError(error);
+  }, []);
 
   const handleCustomButtonClick = useCallback(
     (ev: DailyEventObjectCustomButtonClick) => {
@@ -324,6 +400,7 @@ export default function DailyRoom({ meeting, room }: DailyRoomProps) {
     onLeftMeeting: handleLeave,
     onCustomButtonClick: handleCustomButtonClick,
     onJoinMeeting: handleFrameJoinMeeting,
+    onError: handleError,
   });
 
   useEffect(() => {
@@ -380,13 +457,27 @@ export default function DailyRoom({ meeting, room }: DailyRoomProps) {
   }
 
   if (joinMutation.isError) {
+    const apiDetail = printApiError(
+      joinMutation.error as /*ref 095959E6-01CC-4CF0-B3A9-F65F12F046D3*/ ApiError,
+    );
     return (
       <Center width="100vw" height="100vh">
-        <Text color="red.500">Failed to join meeting. Please try again.</Text>
+        <VStack gap={4}>
+          <Text color="red.500">
+            {apiDetail ?? "Failed to join meeting. Please try again."}
+          </Text>
+          <Button onClick={() => router.push(`/${roomName}`)}>
+            Back to Room
+          </Button>
+        </VStack>
       </Center>
     );
   }
 
+  if (fatalError) {
+    return <FatalErrorScreen error={fatalError} roomName={roomName} />;
+  }
+
   if (!roomUrl) {
     return null;
   }

www/app/lib/apiHooks.ts

@@ -9,6 +9,7 @@ import { MeetingId } from "./types";
 import { NonEmptyString } from "./utils";
 
 /*
+ * ref 095959E6-01CC-4CF0-B3A9-F65F12F046D3
  * XXX error types returned from the hooks are not always correct; declared types are ValidationError but real type could be string or any other
  * this is either a limitation or incorrect usage of Python json schema generator
  * or, limitation or incorrect usage of .d type generator from json schema