[GE users] querying state of consumables

Ross Dickson Ross.Dickson at dal.ca
Wed Nov 14 18:12:50 GMT 2007


    [ The following text is in the "ISO-8859-1" character set. ]
    [ Your display is set for the "ISO-8859-10" character set.  ]
    [ Some special characters may be displayed incorrectly. ]

We have h_vmem configured as consumable on our clusters:

 > qconf -sc | grep h_vmem
h_vmem              h_vmem     MEMORY      <=    YES         YES        
2G       0
 > qconf -se cl002 | grep h_vmem
complex_values        h_vmem=15G

...and so on.  We often get users complaining that their jobs are not 
being scheduled for unknown reasons, and we often track it down to 
insufficient h_vmem.  But the tracking-down is tedious because I don't 
know of a quick way to determine how much vmem is already reserved on a 
given host by the running and suspended jobs.

Is there a command that shows me what the current amount of a consumable 
is consumed on a given node?  Or do I have to write a script that qstats 
for what jobs are running or suspended, which nodes they're on and how 
many slots, how much h_vmem each one has asked for (or defaulted to), 
and add it all up?


-- 
Ross Dickson         HPC Consultant
ACEnet               http://www.ace-net.ca
+1 902 494 6710      Skype: ross.m.dickson

---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe at gridengine.sunsource.net
For additional commands, e-mail: users-help at gridengine.sunsource.net




More information about the gridengine-users mailing list