Intel® DevCloud
Help for those needing help starting or connecting to the Intel® DevCloud
1803 Discussions

PBS: job killed: cput 279480 exceeded limit 270000

Palina
Novice
1,640 Views

My program allocates 224 cores, and CPU limit is 270000, so 270000/224=1205 seconds. So, for a large graph, after it has run 1205 seconds, PBS will kill the task. Is it possible to have more CPU?

=>> PBS: job killed: cput 279480 exceeded limit 270000
Terminated
0 Kudos
7 Replies
SreedeviK_Intel
Moderator
1,597 Views

Hi,

 

Thank you for posting in Intel Communities.

 

Intel DevCloud for oneAPI nodes have a default CPU time limit of 75 hours (270000 seconds). This time can be extended up to 100 hours using the following parameter while requesting for the node:

$ qsub […] -l walltime=24:00:00 -l cput=100:00:00

Eg:

$ qsub -I -l nodes=1:ppn=2 -l walltime=24:00:00 -l cput=100:00:00

SreedeviK_Intel_0-1695744442922.png

 

If this resolves your issue, make sure to accept this as a solution. This would help others with a similar issue. Kindly let us know if you are still facing issues.

 

Regards,

Sreedevi

 

0 Kudos
Palina
Novice
1,591 Views

Hello,

Thank you for the help! I submitted a job so waiting for the output to see if it works.

I was able to find this solution earlier, but there was not much explanation. It seems to be working from qstat output as it seems to use CPU from two nodes. I wanted to know how accurate this is?

#PBS -l select=4:ncpus=56

Palina_0-1695745613998.png

 

0 Kudos
SreedeviK_Intel
Moderator
1,390 Views

Hi,


We are checking on this internally. We will get back to you with an update shortly.


Regards,

Sreedevi


0 Kudos
SreedeviK_Intel
Moderator
1,330 Views

Hi,

 

Sorry, for the delay in getting back to you. 

PBS directory commands in Intel DevCloud for oneAPI allows you to choose upto 2 nodes but will still have the walltime and CPU time limit.

The PBS command you have used (#PBS -l select=4:ncpus=56) is not accurate since you can access only upto 2 nodes. 

For reference:

https://devcloud.intel.com/oneapi/documentation/advanced-queue/ 

 

If you need any additional information, please let us know.

 

Regards,

Sreedevi

 

0 Kudos
SreedeviK_Intel
Moderator
1,259 Views

Hi,


We have not heard back from you. Could you please let us know whether we can go ahead and close this thread?


Regards,

Sreedevi


0 Kudos
SreedeviK_Intel
Moderator
1,198 Views

Hi,


We have not heard back from you. This thread will no longer be monitored by Intel. If you need further assistance, please post a new question.


Regards,

Sreedevi


0 Kudos
Palina
Novice
1,089 Views

Hi SreedeviK,

 

Your solution did help and gave possibility to have more CPU time. It is still not enough for my research analysis but it does allow to create less jobs. I do understand that this is a free resource. Sorry for the late reply I was not able to login to this portal for some reason.

0 Kudos
Reply