Of our four UPSes, one has failed and is currently in bypass, and two have weak batteries.
All three will be replaced with new units.
I will start shutdown early Saturday morning. We don’t know exactlly how much time the electricians need, so we plan for both Saturday and Sunday. It is a good chance that they will finish on Saturday.
All machines and storage (with a couple of exceptions) will be shut down.
07.01.17, 06:35 Shutting down workstations
07.01.17, 08:00 Storage system and servers down
07.01.17, 10:00 UPSes powered off. Electricians working
07.01.17, 14:00 Electricians done. Starting up
07.01.17, 15:30 Starting up storage
07.01.17, 17:25 Storage and servers up. Starting workstations and compute nodes
07.01.17, 19:05 All machines up, except capra, regulus, suhail and vega. They seem to have some hardware problem. Will be debugged on Monday. Done for today.
07.01.17, 21:20 We had a power outage at 20:53. We are without UPS on part of the system until the new UPSes arrive, and owl1, owl2 and owl5 went down.
08.01.17, 11:00 The power outage blew a fuse, leaving the StorNext storage without power redundancy. Fixed. owl1, owl2 and owl5 also rebooted.