You are here
|Post date||System Status:||Update Notes|
|2016-10-27 - 22:29 PDT||Online||
Breezy back to normal operations after system upgrade.
Breezy has been upgraded to the latest CentOS 6.8 operating system.
|2016-10-13 - 02:05 PDT||Conditions||
Wednesday evening, Oct. 12 - Job scheduling restored on Breezy after system upgrade
After a file system problem disrupted service on Breezy on Wednesday morning, scheduling has now been restored, but, on a reduced number of nodes. During the outage, the Linux system and batch system software were updated. The new system will use cpusets, a method of confining the effects of one job on another by more strictly limiting jobs to the resources requested. Some compute nodes still remain to be updated, as running jobs on those nodes will be allowed to complete first. As some jobs were lost due to the file system problem, please check output carefully and resubmit jobs as necessary. Sorry for the inconvenience.
|2016-10-12 - 09:50 PDT||Offline||
Breezy, Lattice, and Parallel are not available due to Filesystem problem
A non disruptive file system upgrade in Calgary did not go according to plan resulting in intermittent access to the file systems.
Some jobs have been lost. We are working with the vendor.
|2016-07-11 - 23:01 PDT||Online||
System fully operational
Finished on July 12, 2016 - 0:00 MDT
|2016-07-08 - 17:22 PDT||Conditions||
The file system that serves Lattice, Parallel, and Breezy has
experienced a condition that forced administrators to reduce the load
on the filesystem temporarily by pausing jobs on a significant portion
of the clusters. Jobs will be resumed when the condition has cleared.
Sorry for the inconvenience.
|2016-06-22 - 19:19 PDT||Online||
Breezy is back to normal operations
Network card was replaced and no further issues have been detected.
|2016-06-21 - 23:19 PDT||Conditions||
Poor Network Performance
Breezy is experiencing a failing network card. It is planned to replace this network card July 22 between 8AM and noon. Console sessions to Breezy will periodically lock up until this problem is resolved.