Node-red随机自动重启问题求助附日志片段

Hey there, let's dig into your Node-RED issue where it's restarting randomly (1-2 times daily) and triggering automatic deployments—starting around 7:15 AM with a deployment notification firing at 7:16 AM. Let's break down the clues and next steps:

1. First, parse the system log snippet you shared #

The line rsyslogd was HUPed tells us the system logging service was reloaded or restarted at 6:25 AM. While this doesn't directly cause Node-RED to restart, it's a sign that something on your Raspberry Pi was making system-level changes around that time—this could be a script, backup tool, or even a system maintenance task that might be linked to your later Node-RED issues.

2. Get Node-RED's own logs (critical for pinpointing restarts) #

System logs only give partial context. You need to check Node-RED's dedicated logs to see why it restarted:

• If you're running Node-RED as a system service, use this command to view real-time logs:

  sudo journalctl -u node-red -f
  

  Or pull historical logs for the timeframe in question:

  sudo journalctl -u node-red --since "2024-05-03 07:00"
  

• If you start Node-RED manually, check the terminal output or the log path specified in your Node-RED settings.

3. Figure out what's triggering the automatic deployments #

That deployment notification firing at 7:16 means a deploy action was triggered—here are the most likely culprits:

• File system changes in your Node-RED directory: Node-RED automatically watches for changes to ~/.node-red (like flows.json, settings.js, or installed nodes). If a script, backup tool, or system process modifies files here, it'll trigger a deploy. Check the modification timestamps of files in this directory to match against your deployment time:

  ls -la ~/.node-red --time-style=+'%Y-%m-%d %H:%M:%S'
  

• Cron jobs (scheduled tasks): Check if there's a timed task running around 7:15 AM that might be touching Node-RED files or restarting the service. View your user's cron jobs with:

  crontab -l
  

  And root's cron jobs with:

  sudo crontab -l
  

• Buggy third-party nodes: If you've recently added new nodes (especially those related to deployment or remote management), they might have a bug that triggers unintended deploys. Try disabling these nodes temporarily to see if the issue stops.

4. Check system-level issues that could kill Node-RED #

Node-RED might be getting restarted by the system itself:

• Resource exhaustion: Check if your Pi ran out of memory or CPU around the restart time. Use htop for real-time monitoring, or if you have sysstat installed, pull historical resource data:

  sar -u -r -s 07:00:00 -e 07:30:00 2024-05-03
  

• OOM Killer (Out-of-Memory): The Linux kernel will kill processes if the system runs out of RAM. Check if Node-RED was a victim:

  dmesg | grep -i oom
  

• Automatic system updates: Some systems run package updates in the early morning, which can restart services. Check your apt history to see if updates happened around your restart time:

  cat /var/log/apt/history.log | grep -i "install\|upgrade"
  

5. Verify Node-RED's configuration #

Double-check your settings.js file for any misconfigured auto-deploy settings. Look for the autoDeploy option or any file-watching configurations that might be set to trigger deploys unexpectedly.

内容的提问来源于stack exchange，提问作者Sean Allen