manifoldcf-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Gustavo Beneitez <gustavo.benei...@gmail.com>
Subject Re: error when running jobs
Date Wed, 24 Oct 2018 09:34:26 GMT
Hi Karl,

we reproduced the issue yesterday and managed to get several logs.
The exact error was:

Error: Unexpected jobqueue status - record id 1540306481078, expecting
active status, saw 12,
So we performed a new search on database and what we got was:
[image: mail1.png]

status G, that means  STATUS_PENDINGPURGATORY
[image: mail2.png]
Please find enclosed a log extraction filtered by this record, on line 53
there is an ERROR message "Unexpected jobqueue status - record id
1540306481078, expecting active status, saw 12". I don't know if that's
what you need or maybe we have also to increase general log level.

Thanks in advance.


El mar., 23 oct. 2018 a las 14:28, Gustavo Beneitez (<
gustavo.beneitez@gmail.com>) escribió:

> Thanks Karl, we are going to make new crawls with that property enable and
> will get back to you.
>
> El mar., 23 oct. 2018 a las 10:09, Karl Wright (<daddywri@gmail.com>)
> escribió:
>
>> Add this to your properties.xml:
>>
>> <property name="org.apache.manifoldcf.diagnostics" value="DEBUG"/>
>>
>> This keeps stuff in memory and dumps a lot to the log as well.
>>
>> I'm afraid that groveling through the logs after a failure to confirm
>> it's the same kind of thing we've seen before takes many hours.  I can only
>> promise to do this when I have the time.
>>
>> Karl
>>
>>
>> On Tue, Oct 23, 2018 at 2:34 AM Gustavo Beneitez <
>> gustavo.beneitez@gmail.com> wrote:
>>
>>> I Karl,
>>>
>>> MySQL. As per config variables:
>>> version  5.7.23-log
>>> version comment MySQL Community Server (GPL)
>>>
>>> which file should I enable logging/debugging?
>>>
>>> Thanks!
>>>
>>> El lun., 22 oct. 2018 a las 21:36, Karl Wright (<daddywri@gmail.com>)
>>> escribió:
>>>
>>>> Hi Gustavo,
>>>>
>>>> I have seen this error before; it is apparently due to the database
>>>> failing to properly gate transactions and behave according to the
>>>> concurrency model selected for a transaction.  We have a debugging setting
>>>> you can configure which logs the needed information so that forensics get
>>>> dumped, and when they do, it's apparent what is happening.
>>>>
>>>> Note well that I have never been able to make this problem appear here,
>>>> so I suspect that the issue is related to network latency or some other
>>>> external factor I cannot easily reproduce.
>>>>
>>>> Just so I know -- what database is this?  The place where we've seen
>>>> this is postgresql; later versions of MySql do not seem to have an issue.
>>>>
>>>> Thanks,
>>>> Karl
>>>>
>>>>
>>>> On Mon, Oct 22, 2018 at 1:44 PM Gustavo Beneitez <
>>>> gustavo.beneitez@gmail.com> wrote:
>>>>
>>>>> Hi Karl,
>>>>>
>>>>> lately we are facing job status problems. After a few minutes the job
>>>>> ends suddenly, always the same way:
>>>>>
>>>>> Error: Unexpected jobqueue status - record id 1539339908660, expecting
>>>>> active status, saw 2
>>>>> Error: Unexpected jobqueue status - record id 1539291541171, expecting
>>>>> active status, saw 2
>>>>> Error: Unexpected jobqueue status - record id 1539294182173, expecting
>>>>> active status, saw 2
>>>>> Error: Unexpected jobqueue status - record id 1539338949797, expecting
>>>>> active status, saw 2
>>>>>
>>>>> I did some investigations and a select to the database after the error
>>>>> appeared and BEFORE rerunning the job:
>>>>>
>>>>> SELECT * FROM `jobqueue` WHERE id = 1539336459053 and jobid =
>>>>> 1539269973731
>>>>>
>>>>>
>>>>> it returned status = 'G'
>>>>>
>>>>>
>>>>> After the run was repeated, it finished OK  and same query returned
>>>>> status = 'C'.
>>>>>
>>>>> I don't understand much of the "active" workers but it seems the item
>>>>> is processed twice. Do you have an idea about what we should investigate?
>>>>>
>>>>>
>>>>> Thanks in advance!
>>>>>
>>>>

Mime
View raw message