Improve error messages in garm log #314
Conversation
maigl
force-pushed
the
improve_error_message
branch
from
8230b40 to
8a31d81
Compare
database/sql/jobs.go
Outdated
| @@ -244,7 +246,8 @@ func (s *sqlDatabase) CreateOrUpdateJob(ctx context.Context, job params.Job) (pa | |||
| if err == nil { | |||
| workflowJob.InstanceID = &instance.ID | |||
| } else { | |||
| slog.With(slog.Any("error", err)).ErrorContext(ctx, "failed to get instance by name") | |||
| // This usually is very normal as not all jobs run on our runners. | |||
| slog.DebugContext(ctx, fmt.Sprintf("failed to get instance by name: %s", job.RunnerName)) | |||
There was a problem hiding this comment.
same as above. The level change to DebugContext is fine.
Co-authored-by: Gabriel <samfiragabriel@gmail.com>
maigl
force-pushed
the
improve_error_message
branch
from
8255c5f to
d6de596
Compare
|
I think we need to bump |
|
I added a specific |
|
Yeah. Let's exclude them for now. I don't think anyone will try to set 2^32 MaxRunners. Anytime soon. We may change the types later to avoid the need to convert, or add some overflow checks. But for now, let's jut exclude them. |
done |
|
Thanks @maigl ! |
finding errors in logs is very important and there are some logs that are
errorsbut seem to be rather normal:ourrunners. This always happens in a scenario where our garm runners are not the only runner setup (and we still get the workflow_job events, e.g. enterprise wide hooks)