* fix auto quality recovery after a downswitch
The downswitch callback armed the upswitch probe, but `triggerDownswitch` reset the stall history right after the callback returned, which disarmed it again. Auto quality stayed on the sub stream until the next chunk boundary no matter how much the connection recovered. The governor now arms the probe itself after the reset.
The chunk boundary effect also ran on mount, so when coverage was already cached it immediately undid the low quality cold start the seed effect had just picked. It now only runs when the chunk actually changes.
* fix recording playback quality switches and silent codec failures
A quality switch changes `bufferLength` a commit before the new source arrives, and since it was a dependency of the hls.js setup effect, the player rebuilt once on the outgoing playlist (jumping back to its original `startPosition`) and again on the new one. The buffer length now updates the running instance's config instead.
A fatal codec error with no lower quality stream to fall back to did nothing at all, so playback just sat there with no message. It now shows the playback failure toast, which is also limited to once per source since hls.js and the video element can both report the same failure.
* retry failed WebRTC probes and skip offline streams
The connectivity probe only ever tried the first go2rtc stream and cached its result for the whole page session, so a single offline camera at the top of the go2rtc config marked WebRTC unreachable for every camera until a reload, as did any brief network hiccup. The probe now starts with the stream being viewed and moves on to the next one when go2rtc reports that it can't open the stream's source. A failed result is only reused for 30 seconds, and it's retried on the next mount or when the page becomes visible again.
* fix two-way talk on cameras with AAC audio
The mic button was enabled whenever WebRTC was globally available, but the live view only switched to the WebRTC player when the stream itself qualified for WebRTC, and AAC playback audio disqualifies it. On most cameras the mic showed as on while nothing was sent. Two-way talk only needs the stream's video to connect since the backchannel is sent, not received, so AAC playback audio no longer blocks it. The mic is also turned off when a stream switch makes talk unavailable.
* keep Frigate+ model references when saving the models section
`/api/config` served a Frigate+ model's path as the resolved `/config/model_cache/<id>` file, and since the models list is saved whole, editing any model in the settings UI wrote that cache path back to the config in place of `plus://<id>`. After a restart the model loaded as a custom model with the default labelmap. The config API now reports the `plus://<id>` reference the model was configured with, and the models section drops the fields the Frigate+ model info supplies (size, tensor, pixel format, dtype, and type) instead of pinning them in the config.
* allow models to share shareable detection hardware
The hardware picker treated every device another model listed as taken, so a second model couldn't pick an Intel GPU or the CPU that the first one already used. The backend only rejects reuse of devices that can't be shared (Coral, MemryX), so the picker now matches that and only marks exclusive units as claimed.
* fix model card state and camera counts in the models editor
Model cards were keyed by index, so deleting one handed its state (such as the selected model source tab) to the card after it. Cards are keyed by scene now, which is unique per model.
The camera count on each card also ignored the backend's fallback to the `all` model, so a camera whose detect scene had no model of its own wasn't counted anywhere. It's counted under `all` now, which also feeds the recommended detector count.
* share a unit's temperature across repeated detector devices
Detector temperatures were matched to units by counting detectors of each type, so a device listed twice to run a second inference process (`hailo:PCIe` and `hailo:PCIe#2`) showed the next unit's temperature, or none at all. Distinct devices are numbered now and repeats share their unit's reading.
* update monitored hardware after a runtime config swap
`swap_runtime_config` rebound the stats emitter to the new config but not its `HardwareStats`, which kept polling hardware for the old config and applied camera updates to the discarded object. It now follows the swap along with its camera update subscriber.
* time out model downloads that never respond
`download_from_url` had no timeout, so a proxy or server that accepted the connection and never answered hung the download forever, including runtime downloads during startup. Connect and read timeouts now fail it like any other download error. The read timeout applies per socket read, so large models still finish.
* resolve segment start times in segment order
A camera stream's cached segments are probed concurrently, and each one chained its start off `last_segment_end` as soon as its own probe finished. When segments backed up in the cache and a later probe finished first, it chained off the wrong segment and the earlier one then moved `last_segment_end` backwards, so rows lost their exact adjacency. Probes still run concurrently, but each segment now waits for the one before it to settle its start before resolving its own.
* plan exports from the same coverage the vod route serves
The vod manifest nulls video-only glitch rows on audio-bearing streams, but exports planned their stream runs from the raw coverage, so a glitch row could produce a mixed-stream file or a 404 that failed the export. `null_audio_glitches` now works out each stream's audio composition itself, and exports go through it like the manifest and its realized timelines do.
An unstaged auto export also paged its playlist and chapters over main whenever main had any rows in range, even when the manifest served the range from sub and main only contributed glitches or slivers at the edges. It now reads the rows of the stream its single run actually uses.
* keep staged export chapters aligned across stream hand-offs
Each staged run of a mixed-stream export is rendered from its own pinned vod playlist, and that playlist's first clip snaps back to the preceding keyframe, so every staged file runs up to a GOP longer than its slice of the merged timeline. Chapters were placed on the merged timeline, so they drifted further from the video at every hand-off. Chapter windows for staged exports are now planned the same way each run's playlist is, carrying that keyframe lead-in into the offsets.
* clarify which hardware units only one model can use
* add e2e tests for shareable hardware and Frigate+ model saves
* only show the path field for a Frigate+ model without an API key
Without `PLUS_API_KEY` the models editor has no Frigate+ tab, so a `plus://` model showed every custom model field. The size, format, type, and labelmap fields are all supplied by the Frigate+ model info and ignored for a Frigate+ model, so editing them did nothing. Only the path is shown now, which still lets the model be switched to a custom one.
* keep a configured input_dtype when saving a Frigate+ model
The backend only overwrites `input_dtype` when the Frigate+ model info supplies `inputDataType`, which older models don't, so a configured dtype still matters for them. Saving the models section was dropping it along with the fields the backend always overwrites.
* probe every go2rtc stream until one isn't refused
The probe stopped after three streams, so with three offline cameras ahead of a working one, WebRTC was marked unreachable everywhere. It only moves past a stream when go2rtc refuses it, which is quick, and a stream that hangs still ends the probe at its timeout, so the cap bought nothing.
* only reuse a failed WebRTC probe for the stream it started from
A failed probe was reused for 30 seconds by every caller, so a camera whose stream timed out kept WebRTC unavailable for the next camera viewed, including one opened while that probe was still running. A pass still counts for every stream since it proves the connection, but a failure is only reused by probes that start from the same stream.
* strip input_dtype from Frigate+ models again
A Frigate+ model's config comes entirely from its model info, and a missing `inputDataType` means the `int` default. Keeping `input_dtype` meant switching from a custom model with `input_dtype: float` to a Frigate+ model carried the stale dtype over, with the field hidden so it couldn't be corrected.
* add a notice registry and System Health tab
Problems Frigate detects on its own (ffmpeg crash loops, stuck detectors, failed model downloads, recordings deleted before their retention period) only ever existed as log lines. This adds a `NoticeRegistry` in the main process backed by two tables, an `update_notice` IPC topic so producers in other processes can reach it through the dispatcher, an admin-only API and websocket topic, and a Health tab that lists them. Kinds declare their own mode, severity, and category in one place: state notices are resolved by their producer, event notices are dismissed by the user.
* treat a prerelease as behind its final release
* fix notices clearing early
* rename menu items and update docs
* don't resolve the update notice on a failed version lookup
* Refactor hardware stats to have consolidated ffmpeg, detector, and enrichments running.
* Cleanup hardware access that is not passed into the container
* Remove network stats from hardware refactor
* Pin ruff
* Add python upgrade fixes
This enables python upgrade checks in ruff to look for deprecated types and patterns. This namely fixes:
- usage of deprecated `Typing` which is now built in
- some specific exceptions which are caught and have new aliases
Some specific UP checks were also ignored as they are stylistic / unimportant and likely to cause bugs
* Remove async blocking calls
Use asyncio.to_thread on two remaining blocking calls to fix hanging event thread loop. Enable this specific rule to block it in the future.
* Use proper logging mechanism
* Correctly format logs
* Raise with context
When raising an exception include the from context to improve debugging
* Cleanup
* resolve global record.export.hwaccel_args to fix phantom camera override
* auto-stop debug replay sessions after 12 hours
* docs tweaks
* add more tips to object classification docs
* tweak language
* Store hwaccel errors with timeout so it can retry
* Add error logs for Intel GPU stats
* add area
---------
Co-authored-by: Nicolas Mowen <nickmowen213@gmail.com>
* embed cpu/mem stats into detectors, cameras, and processes
so history consumers don't need the full cpu_usages dict
* support dot-notation for nested keys
to avoid returning large objects when only specific subfields are needed
* fix setLastUpdated being called inside useMemo
this triggered a setState-during-render warning, so moved to a useEffect
* frontend types
* frontend
hide instead of unmount all graphs - re-rendering is much more expensive and disruptive than the amount of dom memory required
keep track of visited tabs to keep them mounted rather than re-mounting or mounting all tabs
add isActive prop to all charts to re-trigger animation when switching metrics tabs
fix chart data padding bug where the loop used number of series rather than number of data points
fix bug where only a shallow copy of the array was used for mutation
fix missing key prop causing console logs
* add isactive after rebase
* formatting
* skip None values in filtered output for dot notation
* debug replay implementation
* fix masks after dev rebase
* fix squash merge issues
* fix
* fix
* fix
* no need to write debug replay camera to config
* camera and filter button and dropdown
* add filters
* add ability to edit motion and object config for debug replay
* add debug draw overlay to debug replay
* add guard to prevent crash when camera is no longer in camera_states
* fix overflow due to radix absolutely positioned elements
* increase number of messages
* ensure deep_merge replaces existing list values when override is true
* add back button
* add debug replay to explore and review menus
* clean up
* clean up
* update instructions to prevent exposing exception info
* fix typing
* refactor output logic
* refactor with helper function
* move init to function for consistency
* Add rockchip temps
* Add support for GPU and NPU temperatures in the frontend
* Add support for Nvidia temperature
* Improve separation
* Adjust graph scaling
* Add Hailo temperature retrieval
* Refactor `get_hailo_temps()` to use ctxmanager
* Show Hailo temps in system UI
* Move hailo_platform import to get_hailo_temps
* Refactor temperatures calculations to use within detector block
* Adjust webUI to handle new location
---------
Co-authored-by: tigattack <10629864+tigattack@users.noreply.github.com>
* add camera connection quality metrics and indicator
* formatting
* move stall calcs to watchdog
* clean up
* change watchdog to 1s and separately track time for ffmpeg retry_interval
* implement status caching to reduce message volume
* Update version
* Create scaffolding for case management (#21293)
* implement case management for export apis (#21295)
* refactor vainfo to search for first GPU (#21296)
use existing LibvaGpuSelector to pick appropritate libva device
* Case management UI (#21299)
* Refactor export cards to match existing cards in other UI pages
* Show cases separately from exports
* Add proper filtering and display of cases
* Add ability to edit and select cases for exports
* Cleanup typing
* Hide if no unassigned
* Cleanup hiding logic
* fix scrolling
* Improve layout
* Camera connection quality indicator (#21297)
* add camera connection quality metrics and indicator
* formatting
* move stall calcs to watchdog
* clean up
* change watchdog to 1s and separately track time for ffmpeg retry_interval
* implement status caching to reduce message volume
* Export filter UI (#21322)
* Get started on export filters
* implement basic filter
* Implement filtering and adjust api
* Improve filter handling
* Improve navigation
* Cleanup
* handle scrolling
* Refactor temperature reporting for detectors and implement Hailo temp reading (#21395)
* Add Hailo temperature retrieval
* Refactor `get_hailo_temps()` to use ctxmanager
* Show Hailo temps in system UI
* Move hailo_platform import to get_hailo_temps
* Refactor temperatures calculations to use within detector block
* Adjust webUI to handle new location
---------
Co-authored-by: tigattack <10629864+tigattack@users.noreply.github.com>
* Camera-specific hwaccel settings for timelapse exports (correct base) (#21386)
* added hwaccel_args to camera.record.export config struct
* populate camera.record.export.hwaccel_args with a cascade up to camera then global if 'auto'
* use new hwaccel args in export
* added documentation for camera-specific hwaccel export
* fix c/p error
* missed an import
* fleshed out the docs and comments a bit
* ruff lint
* separated out the tips in the doc
* fix documentation
* fix and simplify reference config doc
* Add support for GPU and NPU temperatures (#21495)
* Add rockchip temps
* Add support for GPU and NPU temperatures in the frontend
* Add support for Nvidia temperature
* Improve separation
* Adjust graph scaling
* Exports Improvements (#21521)
* Add images to case folder view
* Add ability to select case in export dialog
* Add to mobile review too
* Add API to handle deleting recordings (#21520)
* Add recording delete API
* Re-organize recordings apis
* Fix import
* Consolidate query types
* Add media sync API endpoint (#21526)
* add media cleanup functions
* add endpoint
* remove scheduled sync recordings from cleanup
* move to utils dir
* tweak import
* remove sync_recordings and add config migrator
* remove sync_recordings
* docs
* remove key
* clean up docs
* docs fix
* docs tweak
* Media sync API refactor and UI (#21542)
* generic job infrastructure
* types and dispatcher changes for jobs
* save data in memory only for completed jobs
* implement media sync job and endpoints
* change logs to debug
* websocket hook and types
* frontend
* i18n
* docs tweaks
* endpoint descriptions
* tweak docs
* use same logging pattern in sync_recordings as the other sync functions (#21625)
* Fix incorrect counting in sync_recordings (#21626)
* Update go2rtc to v1.9.13 (#21648)
Co-authored-by: Eugeny Tulupov <eugeny.tulupov@spirent.com>
* Refactor Time-Lapse Export (#21668)
* refactor time lapse creation to be a separate API call with ability to pass arbitrary ffmpeg args
* Add CPU fallback
* Optimize empty directory cleanup for recordings (#21695)
The previous empty directory cleanup did a full recursive directory
walk, which can be extremely slow. This new implementation only removes
directories which have a chance of being empty due to a recent file
deletion.
* Implement llama.cpp GenAI Provider (#21690)
* Implement llama.cpp GenAI Provider
* Add docs
* Update links
* Fix broken mqtt links
* Fix more broken anchors
* Remove parents in remove_empty_directories (#21726)
The original implementation did a full directory tree walk to find and remove
empty directories, so this implementation should remove the parents as well,
like the original did.
* Implement LLM Chat API with tool calling support (#21731)
* Implement initial tools definiton APIs
* Add initial chat completion API with tool support
* Implement other providers
* Cleanup
* Offline preview image (#21752)
* use latest preview frame for latest image when camera is offline
* remove frame extraction logic
* tests
* frontend
* add description to api endpoint
* Update to ROCm 7.2.0 (#21753)
* Update to ROCm 7.2.0
* ROCm now works properly with JinaV1
* Arcface has compilation error
* Add live context tool to LLM (#21754)
* Add live context tool
* Improve handling of images in request
* Improve prompt caching
* Add networking options for configuring listening ports (#21779)
* feat: add X-Frame-Time when returning snapshot (#21932)
Co-authored-by: Florent MORICONI <170678386+fmcloudconsulting@users.noreply.github.com>
* Improve jsmpeg player websocket handling (#21943)
* improve jsmpeg player websocket handling
prevent websocket console messages from appearing when player is destroyed
* reformat files after ruff upgrade
* Allow API Events to be Detections or Alerts, depending on the Event Label (#21923)
* - API created events will be alerts OR detections, depending on the event label, defaulting to alerts
- Indefinite API events will extend the recording segment until those events are ended
- API event start time is the actual start time, instead of having a pre-buffer of record.event_pre_capture
* Instead of checking for indefinite events on a camera before deciding if we should end the segment, only update last_detection_time and last_alert_time if frame_time is greater, which should have the same effect
* Add the ability to set a pre_capture number of seconds when creating a manual event via the API. Default behavior unchanged
* Remove unnecessary _publish_segment_start() call
* Formatting
* handle last_alert_time or last_detection_time being None when checking them against the frame_time
* comment manual_info["label"].split(": ")[0] for clarity
* ffmpeg Preview Segment Optimization for "high" and "very_high" (#21996)
* Introduce qmax parameter for ffmpeg preview encoding
Added PREVIEW_QMAX_PARAM to control ffmpeg encoding quality.
* formatting
* Fix spacing in qmax parameters for preview quality
* Adapt to new Gemini format
* Fix frame time access
* Remove exceptions
* Cleanup
---------
Co-authored-by: Josh Hawkins <32435876+hawkeye217@users.noreply.github.com>
Co-authored-by: tigattack <10629864+tigattack@users.noreply.github.com>
Co-authored-by: Andrew Roberts <adroberts@gmail.com>
Co-authored-by: Eugeny Tulupov <zhekka3@gmail.com>
Co-authored-by: Eugeny Tulupov <eugeny.tulupov@spirent.com>
Co-authored-by: John Shaw <1753078+johnshaw@users.noreply.github.com>
Co-authored-by: Eric Work <work.eric@gmail.com>
Co-authored-by: FL42 <46161216+fl42@users.noreply.github.com>
Co-authored-by: Florent MORICONI <170678386+fmcloudconsulting@users.noreply.github.com>
Co-authored-by: nulledy <254504350+nulledy@users.noreply.github.com>
* use fallback timeout for opening media source
covers the case where there is no active connection to the go2rtc stream and the camera takes a long time to start
* Add review thumbnail URL to integration docs
* fix weekday starting point on explore when set to monday in UI settings
* only show allowed cameras and groups in camera filter button
* Reset the wizard state after closing with model
* remove footnote about 0.17
* 0.17
* add triggers to note
* add slovak
* Ensure genai client exists
* Correctly catch JSONDecodeError
* clarify docs for none class
* version bump on updating page
* fix ExportRecordingsBody to allow optional name field
fixes https://github.com/blakeblackshear/frigate/discussions/21413 because of https://github.com/blakeblackshear/frigate-hass-integration/pull/1021
* Catch remote protocol error from ollama
---------
Co-authored-by: Nicolas Mowen <nickmowen213@gmail.com>
* Set runtime
* Use count correctly
* Don't assume camera sizes
* Use separate zmq proxy for object detection
* Correct order
* Use forkserver
* Only store PID instead of entire process reference
* Cleanup
* Catch correct errors
* Fix typing
* Remove before_run from process util
The before_run never actually ran because:
You're right to suspect an issue with before_run not being called and a potential deadlock. The way you've implemented the run_wrapper using __getattribute__ for the run method of BaseProcess is a common pitfall in Python's multiprocessing, especially when combined with how multiprocessing.Process works internally.
Here's a breakdown of why before_run isn't being called and why you might be experiencing a deadlock:
The Problem: __getattribute__ and Process Serialization
When you create a multiprocessing.Process object and call start(), the multiprocessing module needs to serialize the process object (or at least enough of it to re-create the process in the new interpreter). It then pickles this serialized object and sends it to the newly spawned process.
The issue with your __getattribute__ implementation for run is that:
run is retrieved during serialization: When multiprocessing tries to pickle your Process object to send to the new process, it will likely access the run attribute. This triggers your __getattribute__ wrapper, which then tries to bind run_wrapper to self.
run_wrapper is bound to the parent process's self: The run_wrapper closure, when created in the parent process, captures the self (the Process instance) from the parent's memory space.
Deserialization creates a new object: In the child process, a new Process object is created by deserializing the pickled data. However, the run_wrapper method that was pickled still holds a reference to the self from the parent process. This is a subtle but critical distinction.
The child's run is not your wrapped run: When the child process starts, it internally calls its own run method. Because of the serialization and deserialization process, the run method that's ultimately executed in the child process is the original multiprocessing.Process.run or the Process.run if you had directly overridden it. Your __getattribute__ magic, which wraps run, isn't correctly applied to the Process object within the child's context.
* Cleanup
* Logging bugfix (#18465)
* use mp Manager to handle logging queues
A Python bug (https://github.com/python/cpython/issues/91555) was preventing logs from the embeddings maintainer process from printing. The bug is fixed in Python 3.14, but a viable workaround is to use the multiprocessing Manager, which better manages mp queues and causes the logging to work correctly.
* consolidate
* fix typing
* Fix typing
* Use global log queue
* Move to using process for logging
* Convert camera tracking to process
* Add more processes
* Finalize process
* Cleanup
* Cleanup typing
* Formatting
* Remove daemon
---------
Co-authored-by: Josh Hawkins <32435876+hawkeye217@users.noreply.github.com>
* Add npu usages as a statistic
* Support showing NPU stats in dashboard
* Add sys volume mount for npu usages
* Fix type
* Simplify check
* Cleanup
* Cleanup
* Actually send result to face registration
* Define postprocessing api and move face processing to fit
* Standardize request handling
* Standardize handling of processors
* Rename processing metrics
* Cleanup
* Standardize object end
* Update to newer formatting
* One more
* One more
* Get stats for embeddings inferences
* cleanup embeddings inferences
* Enable UI for feature metrics
* Change threshold
* Fix check
* Update python for actions
* Set python version
* Ignore type for now
* Subclass Process for audio_process
* Introduce custom mp.Process subclass
In preparation to switch the multiprocessing startup method away from
"fork", we cannot rely on os.fork cloning the log state at fork time.
Instead, we have to set up logging before we run the business logic of
each process.
* Make camera_metrics into a class
* Make ptz_metrics into a class
* Fixed PtzMotionEstimator.ptz_metrics type annotation
* Removed pointless variables
* Do not start audio processor when no audio cameras are configured
