[Neolith-users] Unexpected power-loss yesterday afternoon/evening

[Pär Andersson]

Dear user of Neolith,

Yesterday (Monday June 3rd) NSC suffered an unplanned and complete loss of 
external power. Power went out just before 17.00 and was restored around 35m 
later. All compute nodes of Neolith was emergency stopped so running jobs was 
killed. The servers, including login node and disk servers were kept running 
on UPS power.

After bringing the compute nodes back online we performed a series of large 
scale Infiniband tests over the entire cluster, related to the problems we 
had last week. We also upgraded the System ROM on some of the compute nodes.

After this work had been successfully completed Neolith was put back into 
production around 21.00.

Best regards,
Pär Andersson
NSC