[Triolith-users] Short service stop of Triolith on Friday at 12:00

[Mats Kronberg]

Dear Triolith users,

On Friday at 12:00 CEST we will stop all Triolith compute nodes. This
is due to a required upgrade of the firmware of the core Infiniband
switch in Triolith.

The switch vendor believes this will solve a problem that affects
performance for some jobs.

No running or queued jobs will be affected by this. Instead we have
configured the system to not allow any jobs to start if they can not
finish before Friday 12:00. E.g right now, the longest job that can
start is 43 hours.

As soon as the upgrade is complete and we have verified that the
Infiniband network still works, we will allow queued jobs to start
again. It's hard to say exactly how long this will take, but I
estimate somewhere between 30 minutes and 3 hours.

We will also reboot the login node during this service stop. This will
take 5-15 minutes, but apart from that time, the login node will be
available on Friday, so you can still access your data and submit
jobs.

If you have any questions regarding this, please contact support at nsc.liu.se.

-- 
Mats Kronberg, NSC Support <support at nsc.liu.se>