Too many processes are running, leading to unusable syste,

[gobdevel]

I have deployed wikijs on cloud. After few days it was not able to save any new page. It was giving some unexpected error. Upon inspecting I found the container has used all its resources and was unable to spawn any new process -

Version -
latest (2) - Digest - 01b4e126b4ac

docker logs wiki
2025-11-02T19:14:21.307Z [MASTER] warn: spawn /usr/local/bin/node EAGAIN
2025-11-02T19:14:21.307Z [MASTER] warn: spawn /usr/local/bin/node EAGAIN
2025-11-02T19:16:36.566Z [MASTER] error: Syncing with storage target git: [ FAILED ]
2025-11-02T19:16:36.566Z [MASTER] error: Error: spawn git EAGAIN
    at ChildProcess._handle.onexit (node:internal/child_process:285:19)
    at onErrorNT (node:internal/child_process:483:16)
    at process.processTicksAndRejections (node:internal/process/task_queues:82:21)
2025-11-02T19:20:14.482Z [MASTER] warn: Error: spawn git EAGAIN
    at ChildProcess._handle.onexit (node:internal/child_process:285:19)
    at onErrorNT (node:internal/child_process:483:16)
    at process.processTicksAndRejections (node:internal/process/task_queues:82:21)

homestack-wiki:/wiki$ top
bash: fork: retry: Resource temporarily unavailable
bash: fork: retry: Resource temporarily unavailable
bash: fork: retry: Resource temporarily unavailable
^Cbash: fork: Interrupted system call

homestack-wiki:/wiki$ ulimit -u
unlimited
homestack-wiki:/wiki$ ps
bash: fork: retry: Resource temporarily unavailable
bash: fork: retry: Resource temporarily unavailable
bash: fork: retry: Resource temporarily unavailable

After restarting container, things started working and I am noticing too many git processes is already spawned, looks like the process management need improvement.

homestack-wiki:/wiki$ ps 
PID   USER     TIME  COMMAND
    1 node      0:15 node server
   48 node      0:00 [git]
   58 node      0:00 [git]
  102 node      0:00 [git]
  125 node      0:00 [git]
  132 node      0:00 bash
  160 node      0:00 [git]
  163 node      0:00 [git]

[bikerpatch]

I've seen the same. Deployed to microk8s on RPi, I've seen 312 zombie processes for [git] after a period of 24h, traced back to the wikijs deployment.

[BPowell76]

This sounds like a configuration issue. I run this "bare metal" without containerization and it works fine on a 2 vCPU VPS I am renting.

[fvogl]

I'm also affected with the git zombie processes. Did anyone find a way to avoid them?

[madduck]

I have only ever seen this on Docker, not on bare metal.