This is Lite Plone Theme
You are here: Home System Status

System Status

System related news items.

Please subscribe to our RSS news feed.

Scheduled COSMOS maintenance 06/11/2013 Update

Scheduled maintenance on COSMOS2 and COSMIC, Wednesday 6th November 2013.

We are running some tests on the interconnects on our Intel Xeon Phi cards.

Queues will by drained by 9am Wednesday morning.

****

Due to unforeseen networking issues, Cosmos2 will remain unavailable until tomorrow.

Sorry for any inconvenience caused, we are working with SGI engineers to get this fix ASAP.

Regards,

James

Scheduled COSMOS maintenance 23/10/2013. Finished

Schedule COSMOS maintenance Wednesday, 23 October 2013.

We are planning to test/debug a new environment for Intel XeonPhi co-processors and to update the MOAB/Torque workload management system.

All queues will be drained by 9am Wednesday and remaining jobs purged. The latter is needed prior to the move to a new support database infrastructure.

So the jobs will have to be re-submitted.

Sorry for the inconvenience caused,

Andrey

Scheduled COSMOS maintenance 09/10/2013. Finished

This maintenance is scheduled to replace a failed blade

ak-- 5.40pm, Maintenance is finished and all blades are up and running jobs.

Failed blade replacement on cosmos2 - login nodes (cosmos/universe) remain online as well as the home/scratch storage. Cosmos2 projects areas will be unavailable.

Queues on cosmos2 will be drained by 9am prior to shutdown.

Should be back online by lunchtime.

Sorry for the inconvenience caused,

Andrey

Scheduled COSMOS maintenance 05/10/2013. Finished

COSMOS Systems exceptional maintenance - all systems going offline due to electrical works on site.

ak-- 12am 6 October, cosmos2 was brought up over night and queues restarted now. It took a while, because of one blade intermittent failure.

ak-- 6.20pm, an update -- NFS storage and cosmos/universe/cosmic - are back online. cosmos2 will take a bit longer to finish...

All Cosmos systems - compute and login nodes, file storage and web services - will go offline this Saturday, 5 October 2013.  This is because of the electrical shutdown on site to take place from 3pm till 5pm.

We will take this shutdown as an opportunity to perform certain system maintenance tasks, so all Cosmos systems will go offline at 11am on Saturday and hopefully will be back online by 7pm Saturday.

Sorry for the inconvenience caused.

Andrey

Scheduled COSMOS maintenance 11/09/2013

COSMOS Systems regular maintenance - cosmos2 and cosmic only

cosmos2 and cosmic will be offline for maintenance - preventive DIMM replacement and outstanding OS fixes/patches to be applied. cosmos/universe will continue to be accessible for jobs and file access. cosmos2/projects will be inaccessible though.

Andrey

-- 9.15am maintenance under way