Research Server Outage

Research Server Outage

Nov 4, 2009

It appears that there has been a hardware failure on the main research filer. We have contacted the vendor about getting a replacement. In the meantime research home directories and many of the research system services are unavailable. We hope to have that restored either this evening or first thing in the morning.

Update Nov 5, 2009

The research fileserver (howl) was restored to operation last night.

The network link was suffering heavy packet losses and highly variable packet delays causing extremely slow performance, or no response at all.

The data is safe. The data is also mirrored to another Netapp server and the research home directories are backed up to tape.

So far we have been unable to trace the network issue. A few other hosts exhibit similar network instability, so if you notice any problems, please let me know.