Since yesterday i can’t use ag anymore, it blocks my vps and i need to restart the istance after some time. Looks like there’s a memory leak for AG services
Issue is back, i thought it was solved but today i have it again.
Thank you for the report. This strongly indicates a memory leak triggering an Out-of-Memory (OOM) crash on your VPS.
Could you please try these steps:
- Run dmesg | grep -i “killed” to confirm the OS is terminating the process due to memory limits.
- Limit the service’s RAM (e.g., set MemoryMax=80% in your Systemd file or use --memory in Docker). This ensures only the app crashes/restarts, not the entire VPS.
I confirmed the OOM crash. The command sudo journalctl -k -b -1 shows exactly that the language_server process consumed about 23GB of RAM before being killed by the OS. Log snippet: Out of memory: Killed process 4023829 (language_server) total-vm:33324176kB"
Hi @Herian ,
We apologies for the delay in getting back to you. Could you please let us know if you are still experiencing the issue?
Hello Mrinal,
honestly, after months of quota issues with absolutely lacks of communications, continous bugs (every week there’s a new breaking bug) like model errors and unasble limits, i’m not using AG anymore.
I’m using openai codex in vscode now without any issue.
Thank you
2 Likes