Downtime on some StorNext volumes

We need to replace ageing storage servers and metadata storage in the StorNext storage system. We will do migration from the old to the new servers starting on Tuesday January 30.

Not all StorNext volumes are affected, but our home directory volume, /mn/stornext/u3 will need to be migrated. This will be done in the weekend. There will be two reboots of all StorNext machines*, one on Tuesday morning and one on Sunday.

Schedule for the migration:

Time Mount point Volume Task Status
Tuesday 30/1 morning     Reboot, StorNext machines* Done
Tuesday 30/1 morning /mn/stornext/d7 and d9 ITA4 and PSC1 Unmount Done
Wednesday 31/1 morning /mn/stornext/d7 and d9 ITA4 and PSC1 Remount Done
Wednesday 31/1 morning /mn/stornext/d8 and d11 ITA5 and PSC3 Unmount Done
Thursday 1/2 noon /mn/stornextd8 and d11 ITA5 and PSC3 Remount

Done

Thursday 1/2 noon /mn/stornext/d10 PSC2 Unmount Done
Friday 2/2 evening /mn/stornext/d10 PSC2 Remount Done
Friday 2/2 from 17:00     Logout all local users Done
Friday 2/2 from 17:00 /mn/stornext/u3 u3

Unmount

Done
Sunday 4/2  morning     StorNext software upgrade Done
Sunday 4/2 morning     Reboot, StorNext machines* Done
Sunday 4/2 noon /mn/stornext/u3 u3 Remount Done
Sunday 4/2 afternoon Login enabled     Done

Schedule may change.

* StorNext machines:

algol, arietis, arion, beehive - beehive33-gpu, charybdis1-3, dubhe, eagle-eagle3, electra2, enir, euclid-euclid21, hadar, hercules-hercules15, karbana, mimosa2, owl17-owl28, pleiades, pleiades2, proxima, shaula, tsih, tsih2, viscacha, login.astro.uio.no, all MacOS workstations.

By Torben Leifsen
Published Jan. 27, 2018 9:17 PM - Last modified Feb. 5, 2018 10:20 PM