Soren L. Hansen 96acddbc13 Avoid NPE in nomad/command/job_restart.go (#20049) ...

stopAlloc() checks if an allocation represents a system job like this:
```
  if *alloc.Job.Type == api.JobTypeSystem {
    ...
  }
```

This caused the cli to crash:
```
==> 2024-02-29T08:45:53+01:00: Restarting 2 allocations
    2024-02-29T08:45:54+01:00: Rescheduling allocation "6a9da11a" for group "redacted-group"
panic: runtime error: invalid memory address or nil pointer dereference
[signal SIGSEGV: segmentation violation code=0x2 addr=0x20 pc=0x10686affc]

goroutine 36 [running]:
github.com/hashicorp/nomad/command.(*JobRestartCommand).stopAlloc(0x14000b11040, {0x14000996dc0?, 0x0?})
        github.com/hashicorp/nomad/command/job_restart.go:968 +0x25c
github.com/hashicorp/nomad/command.(*JobRestartCommand).handleAlloc(0x14000b11040, {0x14000996dc0?, 0x0?})
        github.com/hashicorp/nomad/command/job_restart.go:868 +0x34
github.com/hashicorp/nomad/command.(*JobRestartCommand).Run.(*JobRestartCommand).Run.func1.func2()
        github.com/hashicorp/nomad/command/job_restart.go:392 +0x28
github.com/hashicorp/go-multierror.(*Group).Go.func1()
        github.com/hashicorp/go-multierror@v1.1.1/group.go:23 +0x60
created by github.com/hashicorp/go-multierror.(*Group).Go in goroutine 1
        github.com/hashicorp/go-multierror@v1.1.1/group.go:20 +0x84
```

Attaching a debugger revealed that `alloc.Job` was set, but
`alloc.Job.Type` was nil. After guarding the `.Type` check with a
`alloc.Job.Type != nil`, it still crashed. This time, `alloc.Job` was
nil.

I was scrambling to get the job running again, so I didn't have the
opportunity to find out why those values were nil, but this change
ensures the CLI does not crash in these situations.

Fixes #20048

2024-03-01 08:07:28 -06:00

153 B

Raw Permalink Blame History

View Raw