Why Does Frigate Camera Stream Freeze After Hours?
- Camera stream instability over long runs
- FFmpeg/go2rtc not reconnecting cleanly
- Keyframe/timestamp drift over time
Problem Description
One or more Frigate camera streams freeze after running continuously for several hours — the live view shows a static image from hours ago, recordings have gaps, and detection stops. The camera's RTSP session timeout drops the connection, FFmpeg does not reconnect automatically, or the FFmpeg process leaks memory until it crashes.
Why This Happens in Real Homes
A Frigate camera stream that freezes only after hours of running points at a long-run stability problem rather than a config error — the stream works fine initially, then the camera drops or corrupts the connection and FFmpeg/go2rtc doesn't recover cleanly. Keyframe/timestamp drift, a camera firmware bug, or unrecovered network hiccups are typical.
Routing the camera through go2rtc gives more robust reconnection handling, which resolves many long-run freezes, and switching the RTSP transport to TCP avoids UDP packet-loss issues that build up over time. Set a shorter keyframe interval on the camera and update its firmware, since the camera is often the source of the instability. If a specific camera is the culprit, a scheduled nightly reboot of that camera is a pragmatic workaround while you address the root cause.
Symptoms
- Stream freezes after hours
- Camera freezes over time
- Freezes then needs a restart
- Stream stalls after running a while
- Frozen frame after hours
- Intermittent long-run freeze
- Stream dies after uptime
- Needs periodic restart
Recognize these? Here's what usually causes it.
Common Causes
- Camera stream instability over long runs
- FFmpeg/go2rtc not reconnecting cleanly
- Keyframe/timestamp drift over time
- Camera firmware stream bug
- Network hiccup not recovered
- Memory/resource buildup
- Camera dropping the connection
- Transport (UDP) issues over time
Most fixes happen in the first 3 steps.
Do not disable reconnect logic to mask unstable stream behavior.
Tools & Requirements
Step-by-Step Solution
Check for camera RTSP session timeout
Many IP cameras have a built-in RTSP session timeout — the camera drops the RTSP connection after a configurable period (often 1-4 hours) if the client does not send keepalive messages. Check your camera's streaming settings for a 'session timeout' or 'connection timeout' value and increase it to the maximum (or disable it). Also check 'max concurrent connections' — if another viewer (app, NVR, second Frigate instance) opens a connection and hits the limit, the camera drops the oldest connection.
Enable FFmpeg reconnection in Frigate
When the camera stream drops, FFmpeg must reconnect automatically. Set the retry interval in frigate.yml: ffmpeg: retry_interval: 10 (retries every 10 seconds). Also add reconnection input args: -rtsp_transport tcp -stimeout 5000000 (5 second timeout in microseconds). Without these settings, FFmpeg may not automatically reconnect after a stream interruption, leaving the camera feed frozen on the last frame indefinitely.
Switch RTSP transport from UDP to TCP
UDP RTSP transport can drop packets on congested networks, eventually causing the stream decoder to fail and freeze. TCP RTSP transport is more reliable over WiFi and congested networks because it guarantees packet delivery. In frigate.yml: ffmpeg: input_args: preset-rtsp-restream (which includes -rtsp_transport tcp). If you had custom input args without -rtsp_transport tcp: add it. TCP transport uses slightly more bandwidth but significantly reduces stream freezes.
Check for memory leaks in the FFmpeg process
Long-running FFmpeg processes can accumulate memory over time (a known issue with certain FFmpeg builds and camera stream formats). If the Frigate host runs out of memory: the OOM killer terminates FFmpeg processes, causing stream freezes until Frigate restarts the process. Monitor memory usage over 24 hours: watch -n 60 'ps aux | grep ffmpeg'. If FFmpeg processes grow continuously: set up a Frigate cron-based restart (once daily during low-activity hours) as a workaround.
Update camera firmware and Frigate version
Camera firmware updates frequently fix RTSP stability bugs — especially long-running session handling. Check your camera manufacturer's website for firmware updates. Also update Frigate to the latest version: newer Frigate versions include updated FFmpeg builds and go2rtc versions that handle stream recovery better. In Docker: docker pull ghcr.io/blakeblackshear/frigate:stable. In the Home Assistant add-on: check for updates in Settings > Add-ons.
Quick Solutions
Still having issues? This is usually the deeper cause below.
Camera issues that start suddenly almost always trace back to an upload bandwidth drop — run a speed test before assuming hardware failure.
Long-run soak tests catch issues that short functional tests miss.
Live view problems that start suddenly usually trace back to an upload speed drop — the camera itself is fine, the bandwidth path to the cloud isn't.
- Camera stream instability over long runs
- FFmpeg/go2rtc not reconnecting cleanly
- Keyframe/timestamp drift over time
- Camera firmware stream bug
- Network hiccup not recovered
Before you go — try one of these (they fix most cases).
Official Manufacturer Manual
Frigate provides official product documentation through their online manual rather than downloadable PDF. Access setup guides, troubleshooting steps, and product specifications for your Frigate Long-Run Stream Stability.
Source: docs.frigate.video
Need More Help? Frigate Support
Note: The contact information below connects you directly to Frigate's official customer support team, not Trunetto. They can help with warranty claims, device replacements, and advanced technical issues.
Accessories owners commonly pair with Frigate Long-Run Stream Stability.
SANDISK 64GB Extreme microSDXC UHS-I Memory Card with Ada...
SANDISK 128GB Ultra microSDXC UHS-I Memory Card with Adap...
SANDISK 64GB Ultra microSDXC UHS-I Memory Card with Adapt...
As an Amazon Associate we earn from qualifying purchases.