Downtime November 15. - 18
We need some downtime to do system upgrades November 15. - 18 .
The following work will be done:
- Upgrade all linux machines to the latest Red Hat Enterprise 7 version
- Upgrade all Mac workstations running MacOS 11.x.x to 11.6.1
- Upgrade Infiniband drivers and software on all compute nodes
- Mount 4 PB new disk storage on all machines
- Update BIOS and firmware on some compute nodes
15.11.21, 18.00 We are done for today. All hyades and eagle nodes are up. sunflower will need another reboot tomorrow morning for firmware and BIOS updates.
15.11.21, 10:25 We are done with todays upgrades.
15.11.21, 07:00 Linux workstations booted. MacOS in progress.
16.11.21 Pleiades nodes are done
16.11.21 Servers and hyades cluster are done
18.11.21 11:20 All beehive and euclid nodes are done
19.11.21 All hercules and owl nodes are done.
We are finished with all machines
Monday November 15.
- Linux and MacOS workstations (we start the work at 07.00).
Tuesday November 16.
- Update of tsih2 (login.astro.uio.no), tsih and sunflower (NFS and samba server). Will be done early (from 07:00)
- Update of electra2, electra3, mimosa2, viscacha, eagle nodes and hyades nodes.
Note that if you are logged in to login.astro.uio.no you will be logged out and all your processes will be terminated.
Wednesday November 17.
- Update and reboot of owl nodes owl31-37 (we start the work at 07:00)
Thursday November 18.
- Update and reboot of beehive and euclid nodes (we start the work at 07:00). These are many nodes, and it can take a while before they become available.
hercules and pleiades nodes will also be updated, but time will be decided together with the PRITS and La Palma groups.
Unless unexpected things occur, linux workstations will only require a short downtime for a reboot.
Compute nodes require more time, so expect longer downtime. They will become available as the upgrades complete.