>>>>>>
ERROR 2019-07-13T16:20:34,259 (Seeding thread) - Exception tossed:
Unexpected job status: 33
org.apache.manifoldcf.core.interfaces.ManifoldCFException: Unexpected job
status: 33
<<<<<<

This is something I have never seen in any of the ManifoldCF tests ever
done.

I am completely at a lost to explain why you are having tremendous
difficulties with very basic crawls.

Could you do me the following favor: download the latest distribution, and
run the single-process example with NO configuration changes.  Use an
Oracle JDK.  Do your crawl and see if you get the same behavior.

Thanks,
Karl



On Sat, Jul 13, 2019 at 3:56 PM Cihad Guzel <[email protected]> wrote:

> In addition my all job are waiting as "End notification" status now.
>
> Cihad Güzel
>
>
> Cihad Guzel <[email protected]>, 13 Tem 2019 Cmt, 22:27 tarihinde şunu
> yazdı:
>
>> Hi Karl,
>>
>> No, this is different new setup. But I use same version for mfc and
>> database. I am tring new setup for my every testing. I didn't see any
>> repeated or non-repeated error logs like before.
>>
>> Then, I have build the jdbc connector from trunk branch and changed the
>> jdbc-connector.jar with new version. Now, There are non-repeated new error
>> log as follow:
>>
>> ERROR 2019-07-13T16:20:34,259 (Seeding thread) - Exception tossed:
>> Unexpected job status: 33
>> org.apache.manifoldcf.core.interfaces.ManifoldCFException: Unexpected job
>> status: 33
>> at
>> org.apache.manifoldcf.crawler.jobs.JobManager.resetSeedJob(JobManager.java:7934)
>> ~[mcf-pull-agent.jar:?]
>> at
>> org.apache.manifoldcf.crawler.system.SeedingThread.run(SeedingThread.java:242)
>> [mcf-pull-agent.jar:?]
>>
>> Cihad Guzel
>>
>>
>> Karl Wright <[email protected]>, 13 Tem 2019 Cmt, 18:46 tarihinde şunu
>> yazdı:
>>
>>> You previously reported errors of the kind that ManifoldCF throws when
>>> it finds that the database seemingly lost transactional integrity.
>>> My question is whether you are still using the same database setup where
>>> you previously got those errors?
>>>
>>> The ArrayIndexOutOfBounds exception applied to JDBC connector metadata
>>> indexing.  If you are using the JDBC connector with metadata without the
>>> patch then it could explain your problem also.  But you would be seeing
>>> exceptions thrown over and over again in your logs.
>>>
>>> Karl
>>>
>>>
>>> On Sat, Jul 13, 2019 at 11:12 AM Cihad Guzel <[email protected]> wrote:
>>>
>>>> Hi Karl,
>>>>
>>>> If you are talking about is
>>>> https://issues.apache.org/jira/browse/CONNECTORS-1613, my setup
>>>> doesn't include this change because I use mfc 2.12. Are you suggesting I
>>>> use a trunk version?
>>>>
>>>> Cihad Güzel
>>>>
>>>> Karl Wright <[email protected]>, 13 Tem 2019 Cmt, 17:27 tarihinde
>>>> şunu yazdı:
>>>>
>>>>> Is this the same setup where you were getting errors because of
>>>>> inconsistent database states?
>>>>> That could lead to this problem, you know.
>>>>>
>>>>> Karl
>>>>>
>>>>>
>>>>> On Sat, Jul 13, 2019 at 10:14 AM Cihad Guzel <[email protected]>
>>>>> wrote:
>>>>>
>>>>>> Hi Karl,
>>>>>>
>>>>>> I also have a job waiting for 12 days as an "Aborting" status.
>>>>>>
>>>>>> Status: Aborting
>>>>>> Start time: 7/1/19 4:01:46 PM
>>>>>> Documents: 10003
>>>>>> Active: 10003
>>>>>> Processed: 10002
>>>>>>
>>>>>> Cihad Guzel
>>>>>>
>>>>>>
>>>>>> Cihad Guzel <[email protected]>, 13 Tem 2019 Cmt, 16:53 tarihinde
>>>>>> şunu yazdı:
>>>>>>
>>>>>>> Hi Karl,
>>>>>>>
>>>>>>> I tried quick-start single process model. After your suggestion , i
>>>>>>> have tried multiprocess-zk-example for zookeeper-based locking. But I 
>>>>>>> have
>>>>>>> the same problem.
>>>>>>>
>>>>>>> Status: Aborting
>>>>>>> Start time: 7/13/19 3:42:10 PM
>>>>>>> Documents: 10003
>>>>>>> Active: 10003
>>>>>>> Processed: 1021
>>>>>>>
>>>>>>> I'm waiting for over an hour for the jdbc job to stop. I have not
>>>>>>> any error logs in my manifolcf log.
>>>>>>>
>>>>>>> Cihad Güzel
>>>>>>>
>>>>>>> Cihad Güzel
>>>>>>>
>>>>>>>
>>>>>>> Karl Wright <[email protected]>, 8 Tem 2019 Pzt, 13:23 tarihinde
>>>>>>> şunu yazdı:
>>>>>>>
>>>>>>>> Are you using file-based locking?
>>>>>>>> If so, I would suggest strongly migrating to zookeeper-based
>>>>>>>> locking.
>>>>>>>> But if you are using the file-based locking, please execute the
>>>>>>>> "lock clean procedure" as follows:
>>>>>>>>
>>>>>>>> - shut down all manifoldcf processes, including the web UI
>>>>>>>> - run the lock-clean script
>>>>>>>> - start the processes again
>>>>>>>>
>>>>>>>> Thanks,
>>>>>>>> Karl
>>>>>>>>
>>>>>>>>
>>>>>>>> On Mon, Jul 8, 2019 at 5:05 AM Cihad Guzel <[email protected]>
>>>>>>>> wrote:
>>>>>>>>
>>>>>>>>> Hi Karl,
>>>>>>>>>
>>>>>>>>> Nothing. I don't have any error log.
>>>>>>>>>
>>>>>>>>> 8 Tem 2019 Pzt 03:18 tarihinde Karl Wright <[email protected]>
>>>>>>>>> şunu yazdı:
>>>>>>>>>
>>>>>>>>>> Hi Cihad,
>>>>>>>>>>
>>>>>>>>>> What does your manifoldcf log have in it?  Any errors?
>>>>>>>>>>
>>>>>>>>>> Karl
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>> On Sun, Jul 7, 2019 at 3:52 PM Cihad Guzel <[email protected]>
>>>>>>>>>> wrote:
>>>>>>>>>>
>>>>>>>>>>> Hi Karl,
>>>>>>>>>>>
>>>>>>>>>>> I mistakenly wrote "Stopping" instead of "Aborting". My job is
>>>>>>>>>>> waiting as "Aborting" status. I have also the same problem while
>>>>>>>>>>> restarting. I am waiting for 2 days for one job.
>>>>>>>>>>>
>>>>>>>>>>> Regards,
>>>>>>>>>>> Cihad Guzel
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>> Cihad Guzel <[email protected]>, 7 Tem 2019 Paz, 22:42
>>>>>>>>>>> tarihinde şunu yazdı:
>>>>>>>>>>>
>>>>>>>>>>>> Hi Karl,
>>>>>>>>>>>>
>>>>>>>>>>>> I have a few jobs. I stopped all of them but only one job is
>>>>>>>>>>>> waiting as "stopping" status.
>>>>>>>>>>>>
>>>>>>>>>>>> I know that some large jobs is waited long time. But, I have
>>>>>>>>>>>> only 1000 rows on database. So, all of jobs crawled small number of
>>>>>>>>>>>> documents. But , It doesn't make much sense to stay status of 
>>>>>>>>>>>> "stopping"
>>>>>>>>>>>> for a long time. How can I identify a problem?
>>>>>>>>>>>>
>>>>>>>>>>>> Postgresql version: 9.4
>>>>>>>>>>>> Manifoldcf version: 2.12
>>>>>>>>>>>>
>>>>>>>>>>>> Regards,
>>>>>>>>>>>> Cihad Guzel
>>>>>>>>>>>>
>>>>>>>>>>>

Reply via email to