TeslaMate crashing on Google Cloud

[powerswk]

I was able to run TeslaMate on Google Cloud free plan e2-micro, but after 1-2 days :3000 would crash and I'm unable to access the site? Does anyone know the reason to this? It only fixes itself when I stop the VM and run it back on.

[eliluong]

Need logs.

[powerswk]

2025-11-23T00:43:49.482626+00:00 tesla2 google_guest_agent_manager[83157]: Failed to get CPU usage for plugin GuestAgentCorePlugin_20251030.02: error reading stat file: open /proc/53629/stat: no such file or directory
2025-11-22 16:43:49.483 PST
2025-11-23T00:43:49.482859+00:00 tesla2 google_guest_agent_manager[83157]: Failed to get memory usage for plugin Gu

I'm wondering if the e2-micro doesn't have enough resources to handle it.

[eliluong]

I'm not sure either. You would need to SSH into the VM, see if the actual VM still running, and maybe check the Docker logs. Need to figure out if it is Docker issue or VM issue. Teslamate only use a few hundred MB memory, e2-micro with 1GB memory is more than enough. Recommend you to use Gemini or ChatGPT, and it walk you through troubleshooting, etc.

[powerswk]

This is what ChatGPT says

Your Grafana is completely choking on the database.
Every error in your log points to one root cause:

🚨 ROOT PROBLEM: Your SQLite database is locked and Grafana can’t read/write it.

This is why:

Login page returns 500

Cleanup jobs fail

Requests hang for minutes

“database is locked” spams constantly

Session tokens fail (token needs to be rotated)

Grafana can’t load templates (Error rendering template: error)

This happens 100% of the time when Grafana is installed on a tiny VM with a slow disk, especially on Google Cloud VIA Docker.

If TeslaMate or other containers are running heavily on the same machine → the disk I/O gets overwhelmed and SQLite dies.