Hugemem unavailable

Incident Report for NeSI Status

Resolved

This incident has been resolved.

Posted Jun 19, 2019 - 17:11 NZST

Monitoring

The infiniband problem with wbh001 has now been resolved. This hugemem node is now available and running jobs in Mahuika slurm.

Posted Jun 19, 2019 - 17:06 NZST

Investigating

Since the platform maintenance last Thursday (13th June) node wbh001, the only node in the hugemem partition of Mahuika, has been offline due to an Infiniband networking problem. Vendor support has been engaged and severity escalated, however an estimated time to recovery is unavailable as the root cause is still being investigated.

Posted Jun 19, 2019 - 11:10 NZST

This incident affected: NeSI HPC Compute Infrastructure (Mahuika Ancillary nodes).