manifoldcf-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Karl Wright <daddy...@gmail.com>
Subject Re: How to fix: Error: Unexpected jobqueue status
Date Mon, 08 Apr 2013 14:52:45 GMT
I've committed to trunk code that will independently track data
modifications to the jobqueue table which change job statuses.  If there is
a failure, a printout will be dumped to the manifoldcf log which will allow
me to do the necessary forensics to determine the cause of the problem.

The code is, however, not always enabled.  You have to enable it in
properties.xml by setting the value "org.apache.manifoldcf.diagnostics" to
the value "DEBUG".  If you run ManifoldCF in this mode, you should also
expect it to be somewhat slower and use significantly more memory - so keep
an eye out for out-of-memory conditions.

This diagnostic code will ship with the 1.2 release, unless we manage to
find the problem before then.

Thanks,
Karl



On Wed, Apr 3, 2013 at 8:03 AM, Karl Wright <daddywri@gmail.com> wrote:

> I appreciate it.  I am looking for something (anything, really) that is
> out of the ordinary just prior to when the error occurs.  It so far has not
> happened even in my many-hour load tests.  Best hypotheses I can come up
> with so far involves either database transaction integrity issues, or
> (hopefully) a bug which is triggered when there is an indexing failure.  I
> have not found any evidence of the latter in the code, though, and there's
> not enough data to rule anything out at this point.
>
> Karl
>
>
> On Wed, Apr 3, 2013 at 4:43 AM, Christian Hepworth <
> christian.hepworth@york.ac.uk> wrote:
>
>> Hi Karl,
>>
>> Yes, we are seeing this on PostgreSQL. It's never happened to us before,
>> so it's not a common issue.
>>
>> We have been able to restart the job now, we couldn't yesterday. I will
>> have a look in the history for any messages that may be helpful and post
>> them back to you on here.
>>
>> Thanks
>>
>> Christian
>>
>>
>> On 3 April 2013 08:12, Karl Wright <daddywri@gmail.com> wrote:
>>
>>> The reason I don't think this is the correct explanation is because the
>>> error itself indicates that two threads appear to be working on the same
>>> document at the same time.  That's not supposed to be possible, because
>>> there is only one thread that queues documents for processing, and there is
>>> only one thread it can hand the document to.  When the thread that is
>>> supposed to be handling the document finishes with it, it finds that some
>>> other thread has been there first and already updated the document's status
>>> in the database.
>>>
>>> See CONNECTORS-590 for more detail.
>>>
>>> My current theory is that this happens ONLY on MySQL - which is known to
>>> have transactional integrity bugs.  If you are seeing this on PostgreSQL, I
>>> would really want to know that.
>>>
>>> Thanks,
>>> Karl
>>>
>>>
>>>
>>> On Wed, Apr 3, 2013 at 3:02 AM, Karl Wright <daddywri@gmail.com> wrote:
>>>
>>>> So you believe this error occurs due to requeuing of documents when
>>>> there is an output connector error?  Can you provide some evidence for
>>>> that?  I'd love to resolve this problem but heretofore it has been so rare
>>>> as to be almost impossible to chase down.
>>>>
>>>> Karl
>>>>
>>>>
>>>>
>>>> On Wed, Apr 3, 2013 at 12:16 AM, <pankaj.pandey4@wipro.com> wrote:
>>>>
>>>>> Christian,
>>>>>
>>>>> We have also faced similar situation in the past. We figured out that
>>>>> this was happening because of some issue(maybe memory, not sure though)
>>>>> with "Output Connection"(Solr in our case ). Try reducing Throttling
for
>>>>> the for output connection. For us, it's working fine if we set it to
>>>>> 2(default is 10) on DEV server.
>>>>>
>>>>> HTH.
>>>>>
>>>>> Pankaj
>>>>>
>>>>> -----Original Message-----
>>>>> From: Karl Wright [mailto:daddywri@gmail.com]
>>>>> Sent: Tuesday, April 02, 2013 9:45 PM
>>>>> To: user@manifoldcf.apache.org
>>>>> Subject: Re: How to fix: Error: Unexpected jobqueue status
>>>>>
>>>>> Hi Christian,
>>>>>
>>>>> This error should never occur.  If it does it likely indicates a
>>>>> problem with database transactional integrity.  What database are you
>>>>> using?  Would that be MySQL?
>>>>>
>>>>> Karl
>>>>>
>>>>> On Tue, Apr 2, 2013 at 11:05 AM, Christian Hepworth <
>>>>> christian.hepworth@york.ac.uk> wrote:
>>>>> > Hi there,
>>>>> >
>>>>> > We are experiencing the following error from Manifold CF.
>>>>> >
>>>>> > Error: Unexpected jobqueue status - record id 1363876634288,
>>>>> expecting
>>>>> > active status, saw 2
>>>>> >
>>>>> > How do we go about fixing this to allow our job to run again as
>>>>> normal?
>>>>> >
>>>>> > Thanks
>>>>>
>>>>> Please do not print this email unless it is absolutely necessary.
>>>>>
>>>>> The information contained in this electronic message and any
>>>>> attachments to this message are intended for the exclusive use of the
>>>>> addressee(s) and may contain proprietary, confidential or privileged
>>>>> information. If you are not the intended recipient, you should not
>>>>> disseminate, distribute or copy this e-mail. Please notify the sender
>>>>> immediately and destroy all copies of this message and any attachments.
>>>>>
>>>>> WARNING: Computer viruses can be transmitted via email. The recipient
>>>>> should check this email and any attachments for the presence of viruses.
>>>>> The company accepts no liability for any damage caused by any virus
>>>>> transmitted by this email.
>>>>>
>>>>> www.wipro.com
>>>>>
>>>>
>>>>
>>>
>>
>

Mime
View raw message