[Pulp-dev] 500 error during logrotate

[Jonathon Turel]

Hi David,

Sorry for the delay - have been focusing on 6.3 issues..

Anyway, I was not able to reproduce the problem and get my own stack
trace. However, I see that there is one already attached in the BZ:
https://bugzilla.redhat.com/attachment.cgi?id=1351468

It doesn't perfectly match what's in the issue you sent me but it
seems like the repro steps are basically the same. I think those
having a similar root cause is likely. Do you agree, and if so what's
the next step?

Thanks,

Jonathon

On Mon, Jan 8, 2018 at 5:36 PM, David Davis <daviddavis at redhat.com> wrote:
> Jonathon,
>
> I believe your issue is the same as:
>
> https://pulp.plan.io/issues/3129
>
> This also came up during log rotations in apache/httpd. Any way you can get
> the stacktrace to confirm?
>
>
> David
>
> On Mon, Jan 8, 2018 at 11:49 AM, Jonathon Turel <jturel at redhat.com> wrote:
>>
>> Hi team,
>>
>> I'd like some input on this bug:
>> https://bugzilla.redhat.com/show_bug.cgi?id=1512426
>>
>> The httpd error log there shows what is wrong. I would have expected
>> the request to indeed be handled gracefully but that doesn't seem to
>> be the case.
>>
>> Interestingly, I wasn't able to reproduce the problem with a different
>> API: /pulp/api/v2/users/ so maybe this problem is somehow specific to
>> publishing a repo with a distributor ID.
>>
>> We're thinking of putting some across-the-board retry logic into
>> runcible to get around this, but obviously that isn't preferred. This
>> feels like a WSGI or Pulp problem.
>>
>> Thanks for taking a look & let me know if I can provide more details.
>>
>> Jonathon
>>
>> (let me know if there's a better list for this discussion)
>>
>> _______________________________________________
>> Pulp-dev mailing list
>> Pulp-dev at redhat.com
>> https://www.redhat.com/mailman/listinfo/pulp-dev
>
>