Re: [Condor-users] More information from condor_stats

["David Roberts" <David.Roberts@xxxxxxxxx>]

Thanks for your reply.

I did try this last week but the machines then only go into
claimed-suspended. As condor view assumes that any computer in the
claimed (claimed-busy or claimed-suspended) state is doing work I get
the same situation as before (condor view thinks the node is doing work
but it is'nt).  

Would it be sensible just to manually obtain the required data by
running condor_status every so often? i.e. run it every minute and
create my own history file.

Any more ideas greatly appreciated!



David Roberts

Radiotherapy PhD Student
Joint Physics Department
Institute of Cancer Research and Royal Marsden NHS Trust
Downs Road,
Sutton, Surrey
UK, SM2 5PT


>>> tannenba@xxxxxxxxxxx 14/05/2007 05:49:00 >>>
Quick thought re the below:

Assuming you indeed want the local users to trump the Condor jobs, you
could set the SUSPEND expression in condor_config to suspend the Condor
job when the NonCondorLoad crosses a threshold.  Not only would this be
better for you local users, but it would also cause the startd the
change its reported activity - thus perhaps improving your
logging/graphing situation as well.

---
Todd Tannenbaum
University of Wisconsin-Madison
<-- Sent from a Palm Treo 680 phone -->

-----Original Message-----

From:  "David Roberts" <David.Roberts@xxxxxxxxx>
Subj:  [Condor-users] More information from condor_stats
Date:  Sun May 13, 2007 4:41 am
Size:  2K
To:  <condor-users@xxxxxxxxxxx>

I have setup pool logging on our condor cluster and have been using
condor view to create web pages of the cluster useage.  I would
however
like to obtain more information about the cluster.  In particular I
would like to obtain the Condor load averages for all our nodes.  This
information doesnt appear to be logged in the viewhist files (only
state, keyboard and loadavg etc.).  Is there a way to increase the
amount of information stored in the viewhist files so that when i run
condor_stats more information is returned?

Reason for this : Our cluster is setup so that jobs always run on
every
node.  However as the condor jobs run with a low priority, programs
run
by local users on the nodes take over the cpu.  Unfortunately condor
still reports the node as Claimed-Busy (even though the node is busy
doing a local job and node a condor one).  Therefore when the graphs
of
node state are produced the nodes never appear to be busy.  I would
therefore like to produce a graph of idle, UserLoad and CondorLoad so
that the useage of the cluster is more correctly displayed.

Thanks for any suggestions

David Roberts

Radiotherapy PhD Student
Joint Physics Department
Institute of Cancer Research and Royal Marsden NHS Trust
Downs Road,
Sutton, Surrey
UK, SM2 5PT


The Institute of Cancer Research: Royal Cancer Hospital, a charitable
Company Limited by Guarantee, Registered in England under Company No.
534147 with its Registered Office at 123 Old Brompton Road, London SW7
3RP.

This e-mail message is confidential and for use by the addressee only. 
If the message is received by anyone other than the addressee, please
return the message to the sender by replying to it and then delete the
message from your computer and network.
_______________________________________________
Condor-users mailing list
To unsubscribe, send a message to condor-users-request@xxxxxxxxxxx with
a
subject: Unsubscribe
You can also unsubscribe by visiting
https://lists.cs.wisc.edu/mailman/listinfo/condor-users 

The archives can be found at either
https://lists.cs.wisc.edu/archive/condor-users/ 
http://www.opencondor.org/spaces/viewmailarchive.action?key=CONDOR 

_______________________________________________
Condor-users mailing list
To unsubscribe, send a message to condor-users-request@xxxxxxxxxxx with
a
subject: Unsubscribe
You can also unsubscribe by visiting
https://lists.cs.wisc.edu/mailman/listinfo/condor-users 

The archives can be found at either
https://lists.cs.wisc.edu/archive/condor-users/ 
http://www.opencondor.org/spaces/viewmailarchive.action?key=CONDOR

The Institute of Cancer Research: Royal Cancer Hospital, a charitable Company Limited by Guarantee, Registered in England under Company No. 534147 with its Registered Office at 123 Old Brompton Road, London SW7 3RP.

This e-mail message is confidential and for use by the addressee only.  If the message is received by anyone other than the addressee, please return the message to the sender by replying to it and then delete the message from your computer and network.