May 15, 2018
We finished the maintenance on all nodes. We apologize for the inconvenience.
 
----
 

It turned out that there was a problem with the OCTOPUS scheduler which is currently running and maintenance will be carried out on the following schedule to correct this problem.
In order to limit the influence, maintenance is carried out gradually in the node group.
We apologize for any inconvenience.
 

our maintenance plan

May 8(Tue) 10:00 - 12:00

General purpose CPU nodes: 118 nodes
GPU nodes: 18 nodes
Xeon Phi nodes: 22 nodes
Large-Scale Shared-memory nodes: 1 node

-------------------------------------

May 15(Tue) 10:00 - 12:00

General purpose CPU nodes: 118 nodes
GPU nodes: 19 nodes
Xeon Phi nodes: 22 nodes
Large-Scale Shared-memory nodes: 1 node

a infulence by the maintenance

• OCTOPUS

We stop and maintenance compute nodes one half each as above.
You can submit or execute your job during the maintenance, however, queuing time may be longer because valid nodes become half.
 

• Other systems

There is no effect of maintenance.

remarks

• Please note that the job will not run, if you specified a time exceeded the maintenance start time for "elapstim_req".

 





---

Posted : April 27,2018