[Mauiusers] Multiple standing reservations with TASKCOUNT mutually exclude nodes from *all* users

[Chris Samuel]

On Sat, 25 Sep 2004 04:01 am, Wightman wrote:

An update to the one I forgot to send to the list..

> You are correct, our testing was not complete.  However, our testing in
> Moab does show correct behavior (Moab 4.2.0).
>
> Were you testing Moab 4.0.4 or Moab 4.2.0?  The fix was rolled into
> 4.2.0 but not into 4.0.4 (until today).

Moab 4.0.4, I can't see 4.2.0 in the usual places..

Just tried it with moab-4.0.4p8-snap.1096046762, still no luck. :-(

# showres -n | fgrep .0
node084                    User     sque.0.1053946        N/A    1   -00:01:03     4:34:09  Mon Sep 27 15:25:51
node085                    User      sr1.0.1053951        N/A    1   -00:00:11    INFINITE  Mon Sep 27 15:26:43
node085                    User      sr2.0.1053951        N/A    1   -00:00:11    INFINITE  Mon Sep 27 15:26:43
node085                    User     sque.0.1053946        N/A    1   -00:01:03     4:34:09  Mon Sep 27 15:25:51

SRCFG[sr1] PERIOD=INFINITY
SRCFG[sr1] TASKCOUNT=1 FLAGS=SPACEFLEX
SRCFG[sr1] ACCESS=DEDICATEDRESOURCE
SRCFG[sr1] USERLIST=csamuel

SRCFG[sr2] PERIOD=INFINITY
SRCFG[sr2] TASKCOUNT=1 FLAGS=SPACEFLEX
SRCFG[sr2] ACCESS=DEDICATEDRESOURCE
SRCFG[sr2] USERLIST=dbannon

SRCFG[sque] STARTTIME=08:00:00 ENDTIME=20:00:00
SRCFG[sque] PERIOD=DAY DAYS=MON,TUE,WED,THU,FRI
SRCFG[sque] PROCLIMIT<=4 DEPTH=7
SRCFG[sque] HOSTLIST=node084,node085
SRCFG[sque] MAXTIME=00:15:00+*
SRCFG[sque] ACCESS=DEDICATEDRESOURCE

Changing those back to DEDICATED showed that none of them could get onto
the cluster, even though there were nodes free.. :-(

09/27 15:29:59 MReqCreate(218774,SrcRQ,DstRQ,FALSE)
09/27 15:29:59 INFO:     processing node request line '1'
09/27 15:29:59 INFO:     job '218774' loaded:   1   doehme    users   1800       Idle   0 1096262998   [NONE] [NONE] [NONE] >=      0 >=      0 [NONE] 1096262999
09/27 15:29:59 INFO:     137 PBS jobs detected on RM base
09/27 15:29:59 INFO:     jobs detected: 137
09/27 15:29:59 MRMQueueQuery(QCount,EMsg,SC)
09/27 15:29:59 MPBSLoadQueueInfo(base,NULL,SC)
09/27 15:29:59 INFO:     0 PBS jobs detected on RM base
09/27 15:29:59 INFO:     no queues detected
09/27 15:29:59 ALERT:    job sr1.0.1053951 cannot run in any partition
09/27 15:29:59 ALERT:    cannot select 2 procs in partition '[ALL]' for rsv 'sr1.0.1053951'
09/27 15:29:59 ALERT:    cannot create standing reservation 'sr1'
09/27 15:29:59 ALERT:    job sr2.0.1053951 cannot run in any partition
09/27 15:29:59 ALERT:    cannot select 2 procs in partition '[ALL]' for rsv 'sr2.0.1053951'
09/27 15:29:59 ALERT:    cannot create standing reservation 'sr2'
09/27 15:29:59 ALERT:    job sque.0.1053951 cannot run in any partition
09/27 15:29:59 ALERT:    cannot select 4 procs in partition '[ALL]' for rsv 'sque.0.1053951'
09/27 15:29:59 ALERT:    cannot create standing reservation 'sque'
09/27 15:29:59 ALERT:    job sque.1.1053951 cannot run in any partition
09/27 15:29:59 ALERT:    cannot select 4 procs in partition '[ALL]' for rsv 'sque.1.1053951'
09/27 15:29:59 ALERT:    cannot create standing reservation 'sque'
09/27 15:29:59 ALERT:    job sque.2.1053951 cannot run in any partition
09/27 15:29:59 ALERT:    cannot select 4 procs in partition '[ALL]' for rsv 'sque.2.1053951'
09/27 15:29:59 ALERT:    cannot create standing reservation 'sque'
09/27 15:29:59 ALERT:    job sque.3.1053951 cannot run in any partition
09/27 15:29:59 ALERT:    cannot select 4 procs in partition '[ALL]' for rsv 'sque.3.1053951'
09/27 15:29:59 ALERT:    cannot create standing reservation 'sque'
09/27 15:29:59 ALERT:    job sque.4.1053951 cannot run in any partition
09/27 15:29:59 ALERT:    cannot select 4 procs in partition '[ALL]' for rsv 'sque.4.1053951'
09/27 15:29:59 ALERT:    cannot create standing reservation 'sque'

Changing them back to DEDICATEDRESOURCE got them working again, this time with just sr1 and sr2
sharing the same node..

cheers,
Chris
-- 
 Christopher Samuel - (03)9925 4751 - VPAC Systems & Network Admin
 Victorian Partnership for Advanced Computing http://www.vpac.org/
 Bldg 91, 110 Victoria Street, Carlton South, VIC 3053, Australia

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
Url : http://www.supercluster.org/pipermail/mauiusers/attachments/20040927/8f8c0883/attachment.bin