How do I make it never "exit on failure"?

[Merith-TK]

I am trying to use mittnite as an service manager on Termux, and I am running into the issue where code-server can, occasionally, time out when it tries to start, which causes it to count as an "attempt", but I have maxAttempts removed and canFail set to false, and yet it it is defaulting to 3 and true for some reason?

But because of this mittnite keeps exiting, is there a way I can force mittnite to NOT exit on failure?

/home $ mittnite up --api -c .mittnite/conf/
INFO[05-08-2023 02:06:24] looking for configuration files in .mittnite/conf/
INFO[05-08-2023 02:06:24] found config file: .mittnite/conf/code-server.hcl
INFO[05-08-2023 02:06:24] found config file: .mittnite/conf/sshd.hcl
INFO[05-08-2023 02:06:24] generating configuration files
INFO[05-08-2023 02:06:24] probeServer listens on port 9102
INFO[05-08-2023 02:06:24] waiting for probe readiness
INFO[05-08-2023 02:06:24] remote api listens on unix:///var/run/mittnite.sock
INFO[05-08-2023 02:06:25] initialization complete
INFO[05-08-2023 02:06:25] starting job                                  job.name=sshd
INFO[05-08-2023 02:06:25] starting job                                  job.name=code-server
INFO[05-08-2023 02:06:48] handling signal                               signal="child exited"
ERRO[05-08-2023 02:06:48] job exited with error                         error="exit status 1" job.name=code-server
ERRO[05-08-2023 02:06:48] job exited with error                         error="exit status 1" job.name=code-server
INFO[05-08-2023 02:06:48] remaining attempts                            job.maxAttempts=3 job.name=code-server job.usedAttempts=1
INFO[05-08-2023 02:06:48] starting job                                  job.name=code-server
ERRO[05-08-2023 02:07:13] job exited with error                         error="wait: no child processes" job.name=code-server
ERRO[05-08-2023 02:07:13] job exited with error                         error="wait: no child processes" job.name=code-server
INFO[05-08-2023 02:07:13] remaining attempts                            job.maxAttempts=3 job.name=code-server job.usedAttempts=2
INFO[05-08-2023 02:07:13] starting job                                  job.name=code-server
INFO[05-08-2023 02:07:13] reaped child                                  pid=8168 status=256
INFO[05-08-2023 02:07:37] reaped child                                  pid=8357 status=256
ERRO[05-08-2023 02:07:37] job exited with error                         error="wait: no child processes" job.name=code-server
ERRO[05-08-2023 02:07:37] job exited with error                         error="wait: no child processes" job.name=code-server
ERRO[05-08-2023 02:07:37] reached max retries for job code-server; last error: wait: no child processes
FATA[05-08-2023 02:07:37] service runner stopped with error             error="reached max retries for job code-server; last error: wait: no child processes"
INFO[05-08-2023 02:07:37] shutting down remote api
INFO[05-08-2023 02:07:37] sent SIGTERM to job's process group           job.name=sshd

/home $ cat .mittnite/conf/code-server.hcl
job "code-server" {
  command = "/data/data/com.termux/files/usr/bin/code-server"
  args = [""]
  env = []
  canFail = false
  controllable = true
  stdout = "/home/.mittnite/logs/code-server.log"
  stderr = "/home/.mittnite/logs/code-server.err"
}

[martin-helmich]

I think this may be similar to #63 (which also has a proposed implementation in #64); would that satisfy your requirement?

[Merith-TK]

That would be absolutely perfect!