ICHEC mail #26

Posted: 2006-08-08

Dear ICHEC users,


------------------------------------------------------------------------
Unscheduled downtime (Walton and Hamilton)
------------------------------------------------------------------------

At 4.40am this morning, approximately 240 of our cluster nodes rebooted. The pattern of nodes that went down (in contiguous blocks of 10) suggest that one of the 2 input lines suffered a short loss of power. Also, all login nodes went down, as did the Bull (Hamilton). Dual power supplies in critical storage nodes, disk trays and network service nodes prevented even further losses. All compute nodes are back up now.

Obviously all jobs affected by this downtime will be refunded. Apologies for the break in service.

See http://www.ichec.ie/status for up-to-date information.