[Gimle-users] Re: Power Outage at NSC ~ 17.00-17.10

[KentEngström]

kent at nsc.liu.se (Kent Engström) writes:
> There was a campus-wide power outage (or wider) lasting for about 10
> minutes today at Campus Valla.
>
> This was too long for the Gimle compute nodes, and also for some
> of the older generation Accumulus storage servers.
>
> As I am writing this, Gimle is up and running again with the following
> caveats:
>
> - Four Nehalem and two Harpertown nodes are missing. They will probably
>   require service.

Down to one Nehalem and one Hartertown node now. But in the Nehalem
case, as the nodes are packed into four-node enclosures, we need to keep 
four nodes down during trouble-shooting.

> - NFS filesystems /nobackup/rossby{5,7} are missing.

These are now back.

> - Lustre filesystems /nobackup/rossby{9,10,13} and /nobackup/smhid{4,6}
>   are missing.
>
> My colleagues have been working on this tonight and will continue
> tomorrow.

Still working (we are carefull when it comes to these). We hope to have
them available during the day. We will keep you informed.

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