[Bi-users] DONE: Downtime 2015-05-05 13-14 CEST

[Kent Engström]

kent at nsc.liu.se (Kent Engström) writes:
> Dear Bi, Frost and other users,
>
> NETWORK
>
> NSC and SMHI are going to switch the internal connection between NSC
> and SMHI from 2x2x1G to 2x10G. This requires some router reconfiguring
> that will cause downtime for users at SMHI accessing services at NSC.
>
> This has been scheduled for Tuesday May 5, between 13:00 and 14:00 CEST.

That work has been done now, and as far as we know now it went well.

> BI
>
> We also have some work we need to do on Bi (update OS on login nodes and
> system server). So, we will have to reboot login node during the
> downtime.

Done.

> We also need to do some testing we did not have time for before putting
> Bi into production. That testing will mean that there is a risk of
> killing running jobs if things go bad, but we will try to avoid that.

Done, and we do not think we killed anything.

We will start to reboot nodes today too, but that will be between jobs
as you are used to.

> FROST
>
> Finally, Frost will most likely also be down during this time. We may
> need to reboot servers there too, and running jobs are at risk there
> too.

Nothing done on Frost during this downtime.


-- 
Kent Engström, National Supercomputer Centre
kent at nsc.liu.se, +46 13 28 4444