System went down due to thunderstorm generated power spike. We are working on powering back the system.
UPDATE 11:00: System is back again.
System went down due to thunderstorm generated power spike. We are working on powering back the system.
UPDATE 11:00: System is back again.
RSS feed URL has been changed to http://syslog.hpc.uib.no/feed/. Please update your bookmarks.
Looks like after the reboot of the machine on Friday not all data nodes hosting /work picked up proper settings and /work fs is temporary slow. We are working to resolve this issue ASAP.
Update 10:00 This issue is resolved now.
There are issues with the cooling in the machine room. We've contacted building maintenance team and are hoping to bring machine up in few hours.
Update: 14:00 Cooling problems are over. Machine is back online.
Interactive job submission is working again. (qsub -I)
From this night there is a problem with one of our storage systems serving /work, we are looking into the problem.
Update 02.12.2014 13:00: Issue has been remediated, /work should be OK now.
module swap cray-mpich cray-mpich2
module swap cray-libsci cray-libsci/12.2.0
module load craype-barcelona
xtpe-interlagos
- replaced by
craype-interlagos
System maintenance is still ongoing, during the whole day today.
Update 2014.11.25 18:00 Due to unexpected behaviour during update we regret to inform that the maintenance has to be extended. Will will come later with further updates.
Update 2014.11.25 21:27 We have to postpone opening of hexagon due to issues with the scheduling system. We are working tightly with Cray to fix this issue.
Update 2014.11.26 20:33 Issues with the job submission system requires us to delay opening. It well can be that system will not be opened before next week. We try to fix it as soon as possible.
Update 2014.11.27 11:24 The majority of issues were resolved and Hexagon is now available. One of the main remaining issues is interactive job submission, which will be handled during next week, without stopping machine for an extra maintenance.
The scheduled maintenance of /work-common has finished and is available again on Fimm.
Hexagon maintenance has started as planned. Maintenance work is ongoing.
As part of maintenance /work-common will be unmounted, making it additionally unavailable on Fimm.