Diones storage requires another short maintenance break today. Logging in and starting new jobs will not be possible for about 30 minutes, but running jobs will be halted and should resume afterwards.
Storage on the Dione cluster will be unavailable for a while. In the best case is that running jobs will halt and resume after maintenance. That can't be guaranteed, so try to avoid starting jobs that may run at 10 on thursday.
Another break is needed in a few days, more details after this first break.
The move of Dione and other equipment seems to have succeeded without too much problems.
There may be some still unnoticed problems, but it can be used again.
Dione and a lot of other HPC-equipment has been moved to the new site. No serious issues have emerged, but there has been a few delays. We hope that it will be usable again late afternoon on friday.
The datacenter where Dione is placed needs to be emptied before the end of March. This is planned to be done on monday 9.3 and tuesday 10.3 and needs to be done quickly since there is other equipment that need to be moved later.
The whole cluster will be unavailable during this time. The equipment is getting aged so there is also a non-zero risk of something not functioning.
Apologies for the inconvenience and the very quickly scheduled downtime.
Titan cluster is suffering several hardware failures and there's no guarantee that the system will remain operational in the future.
Hence, move your activities to Dione ASAP.
Dione maintenance finished and cluster is ready to use.
Titan maintenance finished and cluster is ready to use.
Titan will be rebooted now.
Dione will be rebooted next week at latest and max runtime is set to few days.
Dione maintenance finished and cluster is ready to use.
However, node di18 is not functioning.