Downtime on storage system 8/2 to 10/2

We plan to do maintenance on all the stornext disk systems and prepare for installation of more storage in the weekend February 8th to 10th.

All machines with direct Fibre connection to storage and all the compute clusters will be shut down. For all other machines, the stornext volumes will be unavailable.

We start shutdown at 16.00 on Friday 8/2. 

You will not be able to log on to the system over the weekend (see below for status and progress). Email will be available as normal. We will announce below, on Slack and by email when the system is ready.

Please log out before 16.00 on Friday.

The following machines will be shut down:

Servers: login.astro.uio.no (tsih2), tsih, electra2, electra3

Linux workstations: enir, algol, shaula, karbana, arietis, dubhe, hadar, arion

Mac workstations: all

Clusters (all nodes): beehive, eagle, euclid, pleiades, owl, hercules, charybdis

Compute: viscacha, mimosa2

The admin storage area will also be unavailble

All other machines may be rebooted if necessary. 

Status:

15.02.19, 14:00 The problems with owl25 - owl28 are solved and the machines are back online. We are still investigating problems with electra3, owl31-35 and eagle4-6 after upgrading to red Hat 7.5.

11.02.19, 12:20 All euclid nodes are ready

11.02.19, 11:10 beehive32-gpu and beehive33-gpu are ready

11.02.19, 11:08 All hercules nodes are ready

11.02.19, 10:45 pleiades-pleiades12 are ready.

Status Sunday evening:

  • All workstations are up
  • euclid21, mimosa2 and viscacha is up
  • The beehive, eagle, charybdis and owl clusters are up
  • Some work remain on the pleiades, euclid and hercules clusters

There are problems with:

  • owl25-28 and owl31 
  • electra3 
  • beehive32-gpu and beehive33-gpu.

These will be looked at Monday morning

10.02.19, 22:20 Reopened for login on login.astro.uio.no

10.02.19, 12:45 Even if you find that some machines are up, don't log in. We are upgrading and will reboot without notice

10.02.19, 09:30 Client upgrades

09.02.19, 19:44 Tape storage configuration done. New servers configured. Upgrading clients.

08.02.19, 23:30 Main StorNext servers upgraded. Some issues with the Infiniband interfaces beeing investigated (solved).

08.02.19, 17:15 Server upgrades in progress

08.02.19, 17:02 Shutdown complete

09.02.19, 22:30 All StorNext servers upgraded. New storage servers installed. Configuration of new tape storage in progress. Client updates in progress. 

 

 

 

 

 

By Torben Leifsen
Published Jan. 29, 2019 5:01 PM - Last modified Feb. 15, 2019 2:23 PM