[Mageia-sysadm] Valstar needs disk replacement...

[nicolas vigier]

On Wed, 07 Sep 2011, Michael Scherer wrote:

> 
> > (if so we could be nice to mdadm and fail the disk before it gets removed)
> > 
> > But we also should replace the failing disc in ecosse, preferrably by a 
> > raid1 (+raid0) setup.  (and why not the same for jonund when "we" are on 
> > site anyway)
> 
> I suspect that this will wait, as we didn't even planned to go there
> ( even if we talked 3 months ago about that ).

I think we can plan a trip to go there in coming weeks. We have a few
things to do there :
 - replace failing disk on ecosse
 - replace failing disk on valstar
 - add second disk on ecosse and jonund (raid1 for system partition, and
   raid0 for builds partition)
 - upgrade from mandriva 2010.2 to mageia on the servers (we can also do
   it remotely, but it can be safer to be there)
And depending on hardware we can have :
 - a backup server
 - an ARM build node
 - a server to host packagers VMs

Maybe we should also start doing backups from valstar to other nodes.