frigate v0.1.0

Automation

Frigate — a complete local NVR with real-time AI object detection for IP cameras, built for Home Assistant. Deployed as a single host-networked Nomad service.

nomad-pack run frigate --registry nomploy

…or one line (add registry + run)

curl -fsSL https://packs.nomploy.com/install.sh | sh -s -- frigate

Needs nomad-pack on PATH. The script only adds the nomploy registry and runs this pack.

Source ↗ Project ↗ ★ 36.1k ⚑ Report an issue

2 tasks http 8971rtsp 8554webrtc 8555 2 volumes image ghcr.io/blakeblackshear/frigate:stable pinned :stable image bumped today

Variables 13

values.hcl

Save as values.hcl, edit, then run:

nomad-pack run frigate -f values.hcl --registry nomploy

# The name of the Nomad job.
job_name = "frigate"

# The Nomad namespace to deploy into.
namespace = "default"

# The datacenters to deploy to.
datacenters = ["*"]

# The Frigate container image. Pin a tag in production.
image = "ghcr.io/blakeblackshear/frigate:stable"

# Host port for the authenticated Frigate web UI.
port = 8971

# Host port for the internal RTSP restream (go2rtc).
rtsp_port = 8554

# Host port for WebRTC live view (TCP and UDP).
webrtc_port = 8555

# Shared-memory size in bytes for camera frame buffers. Increase for many/high-res cameras.
shm_size = 268435456

# tmpfs size in bytes for /tmp/cache (recording segments).
cache_size = 1073741824

# Named volume for the Frigate config and database (/config).
config_volume = "frigate_config"

# Named volume for recordings and snapshots (/media/frigate).
media_volume = "frigate_media"

# Placement constraints. On a nomploy cluster: attribute = "$${meta.nomploy_control_plane}", operator = "=", value = "true".
constraints = []

# The task resources. Object detection is CPU-heavy without a hardware accelerator.
resources = {
    cpu    = 2000
    memory = 2048
  }

Name	Type	Default	Description
job_name	string	"frigate"	The name of the Nomad job.
namespace	string	"default"	The Nomad namespace to deploy into.
datacenters	list	["*"]	The datacenters to deploy to.
image	string	"ghcr.io/blakeblackshear/frigate:stable"	The Frigate container image. Pin a tag in production.
port	number	8971	Host port for the authenticated Frigate web UI.
rtsp_port	number	8554	Host port for the internal RTSP restream (go2rtc).
webrtc_port	number	8555	Host port for WebRTC live view (TCP and UDP).
shm_size	number	268435456	Shared-memory size in bytes for camera frame buffers. Increase for many/high-res cameras.
cache_size	number	1073741824	tmpfs size in bytes for /tmp/cache (recording segments).
config_volume	string	"frigate_config"	Named volume for the Frigate config and database (/config).
media_volume	string	"frigate_media"	Named volume for recordings and snapshots (/media/frigate).
constraints	list	[]	Placement constraints. On a nomploy cluster: attribute = "$${meta.nomploy_control_plane}", operator = "=", value = "true".
resources	object	{ cpu = 2000 memory = 2048 }	The task resources. Object detection is CPU-heavy without a hardware accelerator.

Back up this pack

This pack stores data in 2 Docker named volumes: frigate_configfrigate_media

restic

# Run on the node hosting this pack. Point restic at your repo first:
#   export RESTIC_REPOSITORY="s3:https://<account>.r2.cloudflarestorage.com/<bucket>"
#   export RESTIC_PASSWORD="<repo-password>"
#   export AWS_ACCESS_KEY_ID=<key>  AWS_SECRET_ACCESS_KEY=<secret>
restic backup \
  /var/lib/docker/volumes/frigate_config/_data \
  /var/lib/docker/volumes/frigate_media/_data

rclone (sync to S3/R2)

rclone sync /var/lib/docker/volumes/frigate_config/_data backup:<bucket>/frigate_config
rclone sync /var/lib/docker/volumes/frigate_media/_data backup:<bucket>/frigate_media

Paths assume the default Docker volume location (/var/lib/docker/volumes). Restore by stopping the job, restoring files into the same volume, and re-running the pack.

Readme

frigate

Frigate is a complete, local network video recorder (NVR) with real-time AI object detection for IP cameras. It runs detection on the decoded video stream (with optional Coral/GPU acceleration), records only when it matters, and integrates tightly with Home Assistant. Everything stays on your own hardware — no cloud.

This pack runs Frigate as a single host-networked Nomad service with the shared-memory and tmpfs settings it needs for frame buffering and recording.

Deploy

nomad-pack run frigate --registry=nomploy

Open http://<node-ip>:8971 and create the admin account. Frigate starts with no cameras — edit config.yml in the config_volume to add them, then restart the job. See the config reference.

Configuration

Variable	Default	Description
`image`	`ghcr.io/blakeblackshear/frigate:stable`	Container image (pin a tag in production).
`port`	`8971`	Host port for the authenticated web UI.
`rtsp_port`	`8554`	go2rtc RTSP restream port.
`webrtc_port`	`8555`	WebRTC live-view port (TCP + UDP).
`shm_size`	`268435456` (256 MB)	Shared memory for frame buffers; raise for cameras.
`cache_size`	`1073741824` (1 GB)	tmpfs size for `/tmp/cache`.
`config_volume`	`frigate_config`	Volume for config and the database (`/config`).
`media_volume`	`frigate_media`	Volume for recordings and snapshots.
`resources`	2000 MHz / 2048 MB	CPU and memory (detection is CPU-heavy on CPU).

A prestart init task seeds a minimal config.yml (only if one doesn't already exist, so your edits are preserved) and fixes volume permissions.