I think you are misunderstanding that metric. Frigate does not run detection unless there is motion. Your average detection fps will be dependent on how much motion there is. Also, detection is often run multiple times in multiple parts of the same frame. What really matters is the inference speed. To see where your detection fps would top out, you take 1000/inference_speed.
Iāve just fired up a fresh Hass OS 64 bit RPi 4 with a USB Coral to start playing with Frigate and Iām impressed how far itās come! Iām having issues getting all cameras added though, I have a total of 6 but Iām starting with my TP Link Tapo C200ās and Iām having troubles getting passed 2 cameras at 1920x1080 5fps without the third green screening when I have HW acceleration enabled.
Iāve enabled debug and seen this in the log. Iāll have a look at GPU memory shortly.
watchdog.test INFO : No frames received from test in 20 seconds. Exiting ffmpeg...
watchdog.test INFO : Waiting for ffmpeg to exit gracefully...
watchdog.test INFO : FFmpeg didnt exit. Force killing...
frigate.video INFO : test: ffmpeg sent a broken frame. memoryview assignment: lvalue and rvalue have different structures
frigate.video INFO : test: ffmpeg process is not running. exiting capture thread...
This should be:
logger:
default: debug
1 Like
I updated to the latest version and now the configuration is empty. I know that itās now trying to read it from HA /config directory. Was the update supposed to move the old file to the new location? Iām not seeing anyone else posting about this. Did I have bad luck?
What was the location of the old configuration file? Iām hoping itās still there. Iād hate it if I have to write everything from scratch again.
Unfortunately the changes that the HA devs made to the Supervisor cause the Frigate configuration to completely disappear. This is of course corrected in the latest version by moving the config out of the Supervisorās control. Unless you have a snapshot or backup, itās unrecoverable. Youāll have to recreate it and place it in the HA config directory as frigate.yml.
I know it has already been asked but I tried without success. Is there a chance to implement openalpr for registration plate recognition and face recognitionā¦it might be overkill but seen your work so far itās just amazing and would be a masterpiece and all on one solution for a bunch of people 
Anyone here using Frigate deployed to Kubernetes via helm? I set up a pv/pvc for frigate, and not sure I configured it correctly as doesnāt seem to be using it.
pv config:
apiVersion: v1
kind: PersistentVolume
metadata:
name: pv-frigate
spec:
capacity:
storage: 100Gi
accessModes:
- ReadWriteOnce
persistentVolumeReclaimPolicy: Retain
claimRef:
namespace: default
name: pv-frigate-claim
mountOptions:
- hard
- nfsvers=4.1
nfs:
path: "/volume1/frigate"
server: nasservername
pvc config:
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
name: pv-frigate-claim
spec:
accessModes:
- ReadWriteOnce
resources:
requests:
storage: 100Gi
frigate helm yaml:
persistence:
data:
enabled: true
storageClass: "-"
existingClaim: pv-frigate-claim
subPath: /
accessMode: ReadWriteOnce
size: 100Gi
## Do not delete the pvc upon helm uninstall
skipuninstall: true
Output of describing the pvc:
Name: pv-frigate-claim
Namespace: default
StorageClass: microk8s-hostpath
Status: Bound
Volume: pv-frigate
Labels: <none>
Annotations: control-plane.alpha.kubernetes.io/leader:
{"holderIdentity":"c3832541-6fcf-11eb-999d-62bd07d5689f","leaseDurationSeconds":15,"acquireTime":"2021-02-19T13:02:06Z","renewTime":"2021-...
pv.kubernetes.io/bind-completed: yes
pv.kubernetes.io/bound-by-controller: yes
Finalizers: [kubernetes.io/pvc-protection]
Capacity: 100Gi
Access Modes: RWO
VolumeMode: Filesystem
Used By: frigate-56ffc44c49-8cdqd
Events: <none>
I donāt see it storing any clips on the nfs side when it detects something, so not sure if I am missing something in my settings. Thank you!
EDIT: Deleted/reposted as replied to the wrong person accidentally, didnāt want to mess anyone up.
I do license plate and face recognition already with a couple of other Docker containers and some automations in Home Assistant. Itās not too hard and I actually like having everything separate rather than all together in one container.
How do you guys build a new docker image?
Letās say that I want to try out the 0.8.2 releaseā¦ Seen that people are making a ābuildā ā¦
Iām new to docker so please bear with my stupidnessā¦ 
that sounds interesting! Would you mind sharing this setup? i would be highy interested . I would like using this stuff a lot.
I had the same issue - missed the breaking change and updated before I realised!
I managed to restore mine as I keep snapshot backups on google drive. If you have a snapshot, the config file is in the backup file for your frigate add-on, in the āadd-on.jsonā file.
If you donāt have a snapshot, then I think you are out of luckā¦
So I have spent all day today, between here and redditt and cant figure this out for the life of me. I keep getting this error message:
Starting nginx nginx
...done.
Starting migrations
peewee_migrate INFO : Starting migrations
Starting migrations
peewee_migrate INFO : Starting migrations
There is nothing to migrate
peewee_migrate INFO : There is nothing to migrate
frigate.mqtt INFO : MQTT connected
detector.coral INFO : Starting detection process: 58
frigate.app INFO : Camera processor started for desk: 61
frigate.edgetpu INFO : Attempting to load TPU as usb
frigate.app INFO : Capture process started for desk: 62
Process detector:coral:
frigate.edgetpu INFO : No EdgeTPU detected.
There is nothing to migrate
peewee_migrate INFO : There is nothing to migrate
frigate.mqtt INFO : MQTT connected
detector.coral INFO : Starting detection process: 58
frigate.app INFO : Camera processor started for desk: 61
frigate.edgetpu INFO : Attempting to load TPU as usb
frigate.app INFO : Capture process started for desk: 62
Process detector:coral:
frigate.edgetpu INFO : No EdgeTPU detected.
Traceback (most recent call last):
File "/usr/local/lib/python3.8/dist-packages/tflite_runtime/interpreter.py", line 152, in load_delegate
delegate = Delegate(library, options)
File "/usr/local/lib/python3.8/dist-packages/tflite_runtime/interpreter.py", line 111, in __init__
raise ValueError(capture.message)
ValueError
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "/usr/lib/python3.8/multiprocessing/process.py", line 315, in _bootstrap
self.run()
File "/usr/lib/python3.8/multiprocessing/process.py", line 108, in run
self._target(*self._args, **self._kwargs)
File "/opt/frigate/frigate/edgetpu.py", line 124, in run_detector
object_detector = LocalObjectDetector(tf_device=tf_device, num_threads=num_threads)
File "/opt/frigate/frigate/edgetpu.py", line 63, in __init__
edge_tpu_delegate = load_delegate('libedgetpu.so.1.0', device_config)
File "/usr/local/lib/python3.8/dist-packages/tflite_runtime/interpreter.py", line 154, in load_delegate
raise ValueError('Failed to load delegate from {}\n{}'.format(
ValueError: Failed to load delegate from libedgetpu.so.1.0
frigate.watchdog INFO : Detection appears to have stopped. Restarting frigate
frigate.app INFO : Stopping...
frigate.events INFO : Exiting event processor...
frigate.record INFO : Exiting recording maintenance...
frigate.object_processing INFO : Exiting object processor...
frigate.object_processing INFO : Exiting object processor...
frigate.events INFO : Exiting event cleanup...
frigate.watchdog INFO : Exiting watchdog...
frigate.events INFO : Exiting event cleanup...
frigate.watchdog INFO : Exiting watchdog...
I saw mention of coral but Iām not using coral, this is the very basic config Iām running with
mqtt:
host: 192.xxx.x.xx
# Optional: port (default: shown below)
port: 1883
# Optional: topic prefix (default: shown below)
# WARNING: must be unique if you are running multiple instances
topic_prefix: frigate
# Optional: client id (default: shown below)
# WARNING: must be unique if you are running multiple instances
client_id: frigate
# Optional: user
user: frigate
# Optional: password
# NOTE: Environment variables that begin with 'FRIGATE_' may be referenced in {}.
# eg. password: '{FRIGATE_MQTT_PASSWORD}'
password: 'password'
cameras:
# Required: name of the camera
desk:
# Required: ffmpeg settings for the camera
ffmpeg:
# Required: A list of input streams for the camera. See documentation for more information.
inputs:
# Required: the path to the stream
# NOTE: Environment variables that begin with 'FRIGATE_' may be referenced in {}
- path: rtsp://username:[email protected]/live
# Required: list of roles for this stream. valid values are: detect,record,clips,rtmp
# NOTICE: In addition to assigning the record, clips, and rtmp roles,
# they must also be enabled in the camera config.
roles:
- detect
- rtmp
width: 1280
height: 720
fps: 5
Iām using two wyze cams with the RTSP firmware.
I appreciate any help or direction
You need to set your detector to cpu (depends on number of cpus you have available):
detectors:
cpu1:
type: cpu
cpu2:
type: cpu
1 Like
Any idea what the issue is here? Using HassOS on Rpi4, had some internet/lan connectivity issues recently and not too sure when it started.
Moreover, quite a few times per hour all my integration entities become unavailable for a split second and then restore
Tried to uninstall and reinstall but it still persists
* Starting nginx nginx
...done.
frigate.app INFO : Creating directory: /tmp/cache
frigate.app INFO : Creating tmpfs of size 512m
mount: /tmp/cache: cannot mount tmpfs read-only.
frigate.app ERROR : Failed to create tmpfs, error code: 8192
Starting migrations
peewee_migrate INFO : Starting migrations
peewee DEBUG : ('CREATE TABLE IF NOT EXISTS "migratehistory" ("id" INTEGER NOT NULL PRIMARY KEY, "name" VARCHAR(255) NOT NULL, "migrated_at" DATETIME NOT NULL)', [])
peewee DEBUG : ('SELECT "t1"."id", "t1"."name", "t1"."migrated_at" FROM "migratehistory" AS "t1" ORDER BY "t1"."id"', [])
There is nothing to migrate
peewee_migrate INFO : There is nothing to migrate
frigate.app INFO : Camera processor started for driveway: 41
frigate.app INFO : Camera processor started for garden: 43
frigate.app INFO : Capture process started for driveway: 44
frigate.app INFO : Capture process started for garden: 46
detector.coral INFO : Starting detection process: 38
frigate.edgetpu INFO : Attempting to load TPU as usb
frigate.edgetpu INFO : TPU found
[mov,mp4,m4a,3gp,3g2,mj2 @ 0x5568db3c10] moov atom not found
/media/frigate/recordings/driveway-20210219192817.mp4: Invalid data found when processing input
frigate.record INFO : bad file: driveway-20210219192817.mp4
[mov,mp4,m4a,3gp,3g2,mj2 @ 0x55a2629c10] moov atom not found
/media/frigate/recordings/garden-20210219192818.mp4: Invalid data found when processing input
frigate.record INFO : bad file: garden-20210219192818.mp4
Wow canāt believe it was that simpleā¦thank you very much!
Has anyone been able to get the wyze cams to work? I have tried multiple resolutions and it just comes over as a green screen.
I use Facebox for face recognition (https://www.home-assistant.io/integrations/facebox), it works pretty well. I messed around with OpenALPR a while ago (https://www.home-assistant.io/integrations/openalpr_local/) and couldnāt get it to work, so I switched over to the free API with Plate Recognizer (platerecognizer.com) and have some simple automations that make everything work together with Frigate. Itās even easier with Frigate 0.8.x than it was with earlier versions.
1 Like
I run OpenALPR but separate from HA. I run it in a dedicated VM and have a dedicated camera just for that. Love it!
I found the solution to my issue with helm chart and my pvc not storing data. It looks like it defaulted the pvc to /data which nothing was stored in. I used this with extraVolume and extraVolumeMounts and now persists my frigate data on my pv as i expected: https://github.com/blakeblackshear/blakeshome-charts/pull/7
@blakeblackshear
issue initializing tmpfs.
This is new.
Running Frigate as add-on in homeassistant. Protection Mode = off
* Starting nginx nginx
...done.
frigate.app INFO : Creating directory: /tmp/cache
frigate.app INFO : Creating tmpfs of size 8g
mount: /tmp/cache: cannot mount tmpfs read-only.
You can add tags to an image name when you perform a run command or write a docker-compose file. For example blakeblackshear/frigate:stable-armv7 is the stable image for arm architecture. You can change that to blakeblackshear/frigate:0.8.1-armv7 to use the version 0.8.1 image instead.
You can search images on Docker Hub to find out what tags are available: Docker
I use docker-compose and the stable image. So whenever the stable image changes versions I can just run docker-compose pull frigate to download the latest image. Running docker-compose up -d frigate will then recreate my frigate container using the latest image that was pulled.