KAUST Supercomputing Laboratory Newsletter 20th July

Maintenance Session Tuesday 2nd August

The next maintenance session will be on Tuesday 2nd August from 09:00 until 17:00. There will be no access to the system during this period.

Running watch on squeue

In the KAUST Supercomputing Laboratory Newsletter of 18th May, we requested all Shaheen users to refrain from running watch on squeue. Unfortunately, this instruction went unobserved by some users, placing an unacceptable load on the SLURM scheduler. We have therefore taken measures to prevent usage of the watch command.

Tip of the Week: Check the balance of your allocation

When a project's allocation has expired and/or has an insufficent balance, jobs scripts will not be released to run on Shaheen. Usually, the reason is either AssocMaxJ and/or AssocGrpC, as shown below:

squeue -u $username

      JOBID  USER    ACCOUNT    NAME  ST REASON         START_TIME                  TIME  TIME_LEFT NODES
    1234567  user1   kxxxx      job1  PD AssocMaxJ         N/A                       0:00 1-00:00:00     1
    1234568  user2   kxxxy      job2  PD AssocGrpC         N/A                       0:00 1-00:00:00     2

Before submitting your jobs, check your project allocation, expiration and balance using the "sb" command. Use “groups” command to find out your projects number.

 

username@cdl3:~> sb kxxxx
Project kxxxx: Project Title
PI: PI_NAME

Allocations     Core hours
--------------------------
2016-06-25         3000000
--------------------------
Expired on      2017-10-16
--------------------------
Allocated          3000000
Charged             586450
--------------------------
Balance             286450
--------------------------

 

Follow us on Twitter

Follow all the latest news on HPC within the Supercomputing Lab and at KAUST, on Twitter @KAUST_HPC.

Previous Announcements

http://www.hpc.kaust.edu.sa/announcements/

Previous Tips

http://www.hpc.kaust.edu.sa/tip/