[GE users] shadow master problems

murple andreas.kuntzagk at mdc-berlin.de
Fri Jul 16 14:28:39 BST 2010


Yes, the act_qmaster is updated to contain "login1".

regards, Andreas

rayson wrote:
> Is the "act_qmaster" file updated to contain the new qmaster name??
> 
> You should also reference the Migration of Qmaster to Another Machine
> HOWTO and the Setting Up A Shadow Master In Grid Engine HOWTO:
> 
> http://gridengine.sunsource.net/howto/sge_migrate.html
> http://gridengine.sunsource.net/howto/shadow.html
> 
> Rayson
> 
> 
> 
> On Fri, Jul 16, 2010 at 8:50 AM, murple <andreas.kuntzagk at mdc-berlin.de> wrote:
>> Hi,
>>
>> today we tested shadow master functionality on our cluster. After making
>> $SGE_ROOT/$SGE_CELL/spool shared between the two master nodes (login1 and
>> login2) stopping the qmaster on login2 (the old master) and removing the lock
>> file first seemed to work. login1 noticed the stale heartbeat and started on own
>> qmaster. Unfortunately the exec-nodes did not notice that change and still tried
>> to contact login2.
>> Error messages:
>>
>> main|node001|W|can't register at "qmaster": unable to contact qmaster using port
>> 6444 on host "login2"
>>
>> Our setup is 6.2 with shared $SGE_ROOT/$SGE_CELL/common
>> $SGE_ROOT/$SGE_CELL/spool is only shared between the login/head nodes.
>>
>> regards, Andreas
>>
>> ------------------------------------------------------
>> http://gridengine.sunsource.net/ds/viewMessage.do?dsForumId=38&dsMessageId=268357
>>
>> To unsubscribe from this discussion, e-mail: [users-unsubscribe at gridengine.sunsource.net].
>>
> 
> ------------------------------------------------------
> http://gridengine.sunsource.net/ds/viewMessage.do?dsForumId=38&dsMessageId=268363
> 
> To unsubscribe from this discussion, e-mail: [users-unsubscribe at gridengine.sunsource.net].

------------------------------------------------------
http://gridengine.sunsource.net/ds/viewMessage.do?dsForumId=38&dsMessageId=268367

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



More information about the gridengine-users mailing list