Castor issues on compute nodes closed
Final ticket report This was fixed by building new base images for the virtual machines. We’re still working on figuring out what happened. For some reason, compute nodes already started...
Updated:
November maintenance window closed
The service on Wednesday 7th of November begins at 09:00. The service window affects all systems in various degrees. For Irma and Bianca the queues will be stopped. For all...
Updated:
Issue with castor (filesystem for bianca) closed
We’re currently seeing an issue on one of the nodes providing castor (the file system for bianca). This may lead to failed read/writes and possibly crashed jobs. We’re working to...
Updated:
/sw/data temporarily invisible on crex. closed
/sw/data was temporarily unavailable under that name on crex (the storage system for rackham) for a while today. Fixed at 19:00.
My VASP jobs trigger SIGSEGV closed
We have several reports of VASP jobs killed by SIGSEGV. The problem is believed to be caused by a recently added security fix to limit a certain memory space of...
Updated:
There is currently a problem logging into Bianca closed
There is currently a problem logging into Bianca. The login nodes that gets started during your login by some reason fails to get fully operational. This results in either broken...
Updated:
Quota issues for some projects on crex (rackham) closed
A few projects on crex were subjected to the wrong quota recently. We have worked around the issue and are fixing the underlying data in SUPR.
September maintenance window closed
The service on Wednesday 5th of September starts at 08:50 (10 min earlier than normal). The service window affects all systems in various degrees. For Irma and Bianca the queues...
Updated:
Cooling issues in the UPPMAX compute room closed
The storm that flew passed Uppsala around 15:00 managed to turn off the cooling pumps and we were forced to emergency shutoff the Rackham, Irma, Dis and Bianca As soon...
Updated:
Files may still appear hidden on Bianca closed
We have received reports and seen a few cases from Bianca and Castor that files sadly may still appear as hidden. We are investigating this issue again, and advice users...
Updated:
Memory issues on some rackham nodes closed
We’re seeing issues related to memory on some rackham nodes. This relates to the kernel at times being unable to get the memory it needs. This can show up as...
Updated:
Issues when moving cross quota boundaries within volumes on castor closed
It seems it’s possible to get gluster (the software used to provide the file system service for castor) in a bad state if one try to move something across a...
Connection problem to bianca-sftp.uppmax.uu.se closed
There is currently a problem with the bianca-sftp. You will most likely not be able to connect. Final ticket report There were problems with a script that set incorrect ACLs...
Updated:
The support will be slower between weeks 28-33 closed
If you need help from the UPPMAX support during the summer you may experience increased response time between July and August due to the summer vacations. The staff remaining on-site...
File quota on home directories closed
With our new solution for home directories, we have now implemented file quotas on home directories. In the future there will be a quota of 100 000 inodes (files/folders) in...
Slurm memory handling incorrect for fat nodes closed
It seems slurm (our job scheduler) has changed its behaviour so allocating a fat node through the -C fat or mem256Gb features will not give you access to the extra...
Updated:
Issues with crex (storage for rackham) during midsummer closed
We’re currently (midsummer) having issues with crex (the storage system for rackham) and access may be very slow or possibly result in failed I/Os. Final ticket report This issue persisted...
UPPMAX Cloud network problems closed
The UPPMAX Cloud network provider is having issues with a central switch. At the moment the cloud is unfortunately not reachable. Hopefully this will be fixed very soon.
Updated:
Problems with Lupus closed
A short while ago Irma’s storage system Lupus started misbehaving and shut parts of itself off. We are investigating why this happened. Final ticket report The problem is now solved....
Updated:
UPPMAX was shutdown on Monday at 13:00 CEST due to loss of cooling closed
We are currently having an issue with cooling in the computer hall. If we soon do not make any progress to get the cooling back we will be forced to...
Updated: