[GE users] resetting job number (jobseqnum)

cjf001 john.foley at motorola.com
Wed Jan 13 04:40:18 GMT 2010


Reuti -

thanks for your test and info.

    John


reuti wrote:
> Am 12.01.2010 um 22:39 schrieb rayson:
>
>> If that's the case, then whenever the jobid wraps around after
>> reaching 9,999,999, then the cluster needs to be empty before the
>> users can submit jobs again??
>
> Maybe it's handled differently in this case.
>
> I submitted some jobs with -h, stopped the qmaster, reset the
> jobseqnum, start the qmaster. After some seconds the jobseqnum will
> be overriden with the number of the job with the highest number.
>
> -- Reuti
>
>
>> Last time when I read the code, I think it does not require the
>> cluster to be empty...
>>
>> Rayson
>>
>>
>>
>> On Tue, Jan 12, 2010 at 4:35 PM, reuti<reuti at staff.uni-marburg.de>
>> wrote:
>>> Am 12.01.2010 um 22:32 schrieb cjf001:
>>>
>>>> Oh yes - lots and lots of them......  I take it that you
>>>> think the system must be "empty" for that to work ?
>>>
>>> Yes, that's my experience.
>>>
>>> -- Reuti
>>>
>>>
>>>>     Thanks,
>>>>
>>>>        John
>>>>
>>>>
>>>> reuti wrote:
>>>>> Hi,
>>>>>
>>>>> Am 12.01.2010 um 21:00 schrieb cjf001:
>>>>>
>>>>>> Guys -
>>>>>>
>>>>>> I'm looking for a way to reset the job numbers back to "1" or
>>>>>> something
>>>>>> low. I read this webpage:
>>>>>
>>>>> was there any running or waiting job in the system?
>>>>>
>>>>> -- Reuti
>>>>>
>>>>>
>>>>>> http://gridengine.info/2007/08/24/reset-grid-engine-job-id-counter
>>>>>>
>>>>>> which says this:
>>>>>>
>>>>>>> Reset Grid Engine Job ID Counter
>>>>>>> Posted by chris on Friday, August 24, 2007
>>>>>>>
>>>>>>> Update 12/2009: It has been pointed out that the actual rollover
>>>>>>> value for the SGE JOB ID is 9,999,999.
>>>>>>>
>>>>>>> In a recent post, Sathish asks:
>>>>>>>
>>>>>>>
>>>>>>> My Current scenario: The job-ids crossed 5000. I'm quite aware
>>>>>>> that i can go with the job-id's till 999999. My Expectation: Is
>>>>>>> there any option to reset the existing job-id's such that the
>>>>>>> next
>>>>>>> job i submit will follow from 1 to ....
>>>>>>> Reuti is quick to mention that the Job ID counter is kept in a
>>>>>>> plaintext file at location:
>>>>>>>
>>>>>>>
>>>>>>> $SGE_ROOT/default/spool/qmaster/jobseqnumber... this does require
>>>>>>> a restart of Grid Engine to take effect.
>>>>>>>
>>>>>>> In another reply, Rayson offers a hint to people interested in
>>>>>>> altering the default value at which the SGE Job ID counter is
>>>>>>> reset back to 0. Apparently this value is encoded as "MAX_SEQNUM"
>>>>>>> in the SGE source code.
>>>>>>>
>>>>>>
>>>>>> However, it doesn't seem to work.  I stop the sgemaster process on
>>>>>> the qmaster,
>>>>>> edit the $SGE_ROOT/default/spool/qmaster/jobseqnumber file (this
>>>>>> should actually
>>>>>> say "$SGE_ROOT/$SGE_CELL/spool/qmaster/jobseqnum" in that webpage,
>>>>>> I think), and
>>>>>> restart the sgemaster process - but after a few seconds that file
>>>>>> gets changed
>>>>>> back to  the large number that it contained before I editted it.
>>>>>> Any ideas
>>>>>> on how to do this, or what I'm missing ?
>>>>>>
>>>>>>       Thanks,
>>>>>>
>>>>>>           John
>>>>>>
>>>>>> ------------------------------------------------------
>>>>>> http://gridengine.sunsource.net/ds/viewMessage.do?
>>>>>> dsForumId=38&dsMessageId=238386
>>>>>>
>>>>>> To unsubscribe from this discussion, e-mail: [users-
>>>>>> unsubscribe at gridengine.sunsource.net].
>>>>>
>>>>> ------------------------------------------------------
>>>>> http://gridengine.sunsource.net/ds/viewMessage.do?
>>>>> dsForumId=38&dsMessageId=238396
>>>>>
>>>>> To unsubscribe from this discussion, e-mail: [users-
>>>>> unsubscribe at gridengine.sunsource.net].
>>>>
>>>> ------------------------------------------------------
>>>> http://gridengine.sunsource.net/ds/viewMessage.do?
>>>> dsForumId=38&dsMessageId=238404
>>>>
>>>> To unsubscribe from this discussion, e-mail: [users-
>>>> unsubscribe at gridengine.sunsource.net].
>>>
>>> ------------------------------------------------------
>>> http://gridengine.sunsource.net/ds/viewMessage.do?
>>> dsForumId=38&dsMessageId=238405
>>>
>>> To unsubscribe from this discussion, e-mail: [users-
>>> unsubscribe at gridengine.sunsource.net].
>>>
>>
>> ------------------------------------------------------
>> http://gridengine.sunsource.net/ds/viewMessage.do?
>> dsForumId=38&dsMessageId=238406
>>
>> To unsubscribe from this discussion, e-mail: [users-
>> unsubscribe at gridengine.sunsource.net].
>>
>
> ------------------------------------------------------
> http://gridengine.sunsource.net/ds/viewMessage.do?dsForumId=38&dsMessageId=238413
>
> To unsubscribe from this discussion, e-mail: [users-unsubscribe at gridengine.sunsource.net].

------------------------------------------------------
http://gridengine.sunsource.net/ds/viewMessage.do?dsForumId=38&dsMessageId=238476

To unsubscribe from this discussion, e-mail: [users-unsubscribe at gridengine.sunsource.net].



More information about the gridengine-users mailing list