Job not shutting down, document status Scheduled field set to a date in 1969

["Mcguinness, Cathal" <Ca...@fmr.com.INVALID>]

Hi,
I am having some difficulty troubleshooting an issue with MCF.
I have 3 jobs which are on a schedule and should take approximately 20 mins at most to complete.
However after 2 days their status is showing as running. I have checked the MCF application logs and do not see any errors or exceptions.
I have also checked the postgres logs for any exceptions and don't see any there either.
The postgres process was restarted some time before the jobs schedule, so not sure if this is a contributing factor. Is there a need to reset the seeding on these jobs, clear history after a postgres restart?
I have noticed in the document status that documents associated with the jobs in question have a status of  "Ready for processing" and a schedule for "12-31-1969 19:00:00" . How is this schedule date set?

Any help or guidance would be greatly appreciated.

Regards
Cathal.

Re: Job not shutting down, document status Scheduled field set to a date in 1969

[Karl Wright <da...@gmail.com>]

Hi Cathal,

The date in question is "start of epoch", which is the earliest date in
milliseconds that can be represented.  There is no implication other than
the document is ready to be processed and is not waiting.

If you restarted postgresql without shutting down ManifoldCF first, it is
conceivable that there are stuck threads as a result of that.  The code
should be resilient against such things but that failure mode is not
exercised very much.  Can you post a thread dump of your agents process?

More to the point -- what kind of synchronization are you using?  Is this a
multi-process installation?  If so, is this zookeeper or file synch?

Karl


On Thu, Aug 13, 2015 at 4:18 AM, Mcguinness, Cathal <
Cathal.Mcguinness@fmr.com.invalid> wrote:

> Hi,
> I am having some difficulty troubleshooting an issue with MCF.
> I have 3 jobs which are on a schedule and should take approximately 20
> mins at most to complete.
> However after 2 days their status is showing as running. I have checked
> the MCF application logs and do not see any errors or exceptions.
> I have also checked the postgres logs for any exceptions and don't see any
> there either.
> The postgres process was restarted some time before the jobs schedule, so
> not sure if this is a contributing factor. Is there a need to reset the
> seeding on these jobs, clear history after a postgres restart?
> I have noticed in the document status that documents associated with the
> jobs in question have a status of  "Ready for processing" and a schedule
> for "12-31-1969 19:00:00" . How is this schedule date set?
>
> Any help or guidance would be greatly appreciated.
>
> Regards
> Cathal.
>
>