Fw: [Mauiusers] question on maui 3.2.6p20: can not get job list from WIKI RM

Hien Nguyen hien1 at us.ibm.com
Wed Nov 5 10:02:43 MST 2008


The contents of my mail was cut off on mauiuser mailing list. Sent again.

Regards,

 Hien Nguyen
Linux Technology Center (Austin)
 Phone: (512) 838-4140            Tie Line: 678-4140
 e-mail: hien1 at us.ibm.com

----- Forwarded by Hien Nguyen/Austin/IBM on 11/05/2008 11:01 AM -----

Hien Nguyen/Austin/IBM
11/04/2008 01:53 PM

To
Eygene Ryabinkin <rea+maui at grid.kiae.ru>
cc
mauiusers at supercluster.org
Subject
Re: [Mauiusers] question on maui 3.2.6p20: can not get job list from WIKI 
RM





Eygene, good day,
Thanks for your help.

I modified maui.cfg to have RMCFG[p6ihopenhpc-ib-3] TYPE=WIKI PORT=7321 
HOST=p6ihopenhpc-ib-3 AUTHTYPE=CHECKSUM
Howver, I got warnings on MRMClusterQuery() and MRMWorkloadQuery() as 
following:

11/04 12:46:32 MSysUpdateTime()
11/04 12:46:32 INFO:     starting iteration 15
11/04 12:46:32 MRMGetInfo()
11/04 12:46:32 MClusterClearUsage()
11/04 12:46:32 MRMClusterQuery()
11/04 12:46:32 WARNING:  no resources detected
11/04 12:46:32 MRMWorkloadQuery()
11/04 12:46:32 WARNING:  no workload detected
11/04 12:46:32 MStatClearUsage(node,Active)
11/04 12:46:32 MClusterUpdateNodeState()

Am I missing something else ?

Regards,

 Hien Nguyen
Linux Technology Center (Austin)
 Phone: (512) 838-4140            Tie Line: 678-4140
 e-mail: hien1 at us.ibm.com




Eygene Ryabinkin <rea+maui at grid.kiae.ru> 
11/04/2008 11:01 AM

To
Hien Nguyen/Austin/IBM at IBMUS
cc
mauiusers at supercluster.org
Subject
Re: [Mauiusers] question on maui 3.2.6p20: can not get job list from WIKI 
RM






Hien, good day.

Tue, Nov 04, 2008 at 08:19:52AM -0600, Hien Nguyen wrote:
> I run maui and slurm 1.3.6 . I found that in maui log there are errors 
and 
> alerts:
> 11/03 23:56:40 ERROR:    command 'CMD=GETNODES ARG=0:ALL'  SC: -300 
> response: 'NONE'
> 11/03 23:56:40 ALERT:    cannot get node list from WIKI RM
> 11/03 23:56:40 ALERT:    cannot load cluster resources on RM (RM 
> 'p6ihopenhpc-ib-3' failed in function 'clusterquery')
> 11/03 23:56:40 WARNING:  no resources detected
> 
> Can someone tell what's wrong with the config of maui and slurm?
> 
> file maui.cfg:
> -------------------------------------
> # maui.cfg 3.2.6p20
> 
> SERVERHOST            p6ihopenhpc-ib-3
> # primary admin must be first in list
> ADMIN1                root
> 
> # Resource Manager Definition
> 
> RMCFG[p6ihopenhpc-ib-3] TYPE=WIKI
> RMPORT          7321
> RMHOST          p6ihopenhpc-ib-3
> RMAUTHTYPE[p6ihopenhpc-ib-3] MUNGE
  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

I very much doubt that Maui support Munge authentication.  You will want
to use 'RMCFG[<host>] TYPE=WIKI PORT=7321 HOST=<host> AUTHTYPE=CHECKSUM'
along with the Slurm's wiki.conf carrying the appropriate 'AuthKey'
directive.  The key itself should contain only digits, it shouldn't
be bigger than 2^32 and the key should be the same as one was used
during Maui compilation (parameter '--with-key' to the configure script).

And you will need the patch mentioned in the list message
  
http://www.clusterresources.com/pipermail/mauiusers/2008-October/003564.html

or to use maui-3.2.6p21-snap.1224706197 that was already patched by
Brian Christiansen.
-- 
Eygene Ryabinkin, Russian Research Centre "Kurchatov Institute"
[attachment "attqcjzu.dat" deleted by Hien Nguyen/Austin/IBM] 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.supercluster.org/pipermail/mauiusers/attachments/20081105/0dd9a738/attachment.html


More information about the mauiusers mailing list