Difference between revisions of "Leonhard Open maintenance (Dec 2020)"

From ScientificComputing
Jump to: navigation, search
Line 1: Line 1:
Due to a network maintenance (replacement of central Ethernet switches in the LEE data center) on '''2-3 December 2020''', the Leonhard Open cluster needs to be '''completely shutdown'''. During this downtime we will also upgrade the firmware of the Lustre file system in Leonhard Open. We are therefore announcing that
+
Due to a major network maintenance in the computer room where Leonhard is located,
  
  The Leonhard Open cluster will be '''OFFLINE''' from '''12:00 pm (noon) on Wednesday, 2 December 2020'''.
+
  this cluster will be '''OFF-LINE'' from '''08:00 AM, Wednesday 2 December''' to '''08:00 AM, Friday 4 December'''.
  
We expect to be again online in the morning of 4 December 2020.
+
During this downtime we will also upgrade the firmware of the Lustre file system in Leonhard Open. As usual, batch queues will be inactivated before the maintenance to ensure that no job gets killed when the cluster is shutdown. Shorter jobs can continue to run if they can finish by the time the cluster is taken offline. There is no action required on your part. This page will be updated before and during this maintenance.
 
 
As usual, batch queues will be inactivated before the maintenance to ensure that no job gets killed when the cluster is shutdown. Shorter jobs can continue to run if they can finish by the time the cluster is taken offline.
 
 
 
There is no action required on your part. This page will be updated before and during this maintenance.
 
  
 
We apologize for any inconvenience this may cause.
 
We apologize for any inconvenience this may cause.

Revision as of 07:55, 12 November 2020

Due to a major network maintenance in the computer room where Leonhard is located,

this cluster will be OFF-LINE from 08:00 AM, Wednesday 2 December' to 08:00 AM, Friday 4 December.

During this downtime we will also upgrade the firmware of the Lustre file system in Leonhard Open. As usual, batch queues will be inactivated before the maintenance to ensure that no job gets killed when the cluster is shutdown. Shorter jobs can continue to run if they can finish by the time the cluster is taken offline. There is no action required on your part. This page will be updated before and during this maintenance.

We apologize for any inconvenience this may cause.