The scratch disk mounted on /scratch on macomp01 and accessible cluster-wide as /scratchcomp01 has become inaccessible sometime between the afternoon of May 24th and Thursday morning, May 26th. This may be a temporary problem resolvable through a reboot so macomp01 has been taken offline from the Torque/Maui queueing system and is not accepting any new jobs although the existing jobs running on it will be allowed to run to completion. Once it is free of running jobs, it will be rebooted.
If a reboot does not resolve this problem, the disk is almost certainly faulty and will have to be replaced. This is a good time to remind all users that the scratch disks on the compute cluster machines are literally just that - meant for temporary storage and not for long term retention of important data since none of the scratch disks are backed up.
Silo upgrade
silo was substantially upgraded at the weekend is back up & online. The documentation is currently in the process of being revised as some things have changed but for most users, nothing has changed although you will probably have to delete the existing ssh host key for silo from your ~/.ssh/known_hosts file before connecting to it for the first time using scp, sftp or ssh.