I guess this one is due for an update 🙂
Last week some helpful engineers came down to try and work out why our servers were misbehaving as they were unable to reproduce the problem in their lab.
After a bit of work, they were able to reproduce the bug on the lab machine as well as our local. It would appear that there is a bug in the LifeCycle Controller Firmware version 126.96.36.1991. which only manifests itself when more than one power supply has power applied to it at the time of boot. The fault does not appear if only one power supply is powered on at boot.
Quite an interesting albeit annoying bug for very few people. For possibly 99% of people out there, they either won’t be using the unified server configurator for deployment, won’t have version 188.8.131.521 of LCC firmware installed, find the ‘lag’ an annoyance but not a showstopper, don’t have a R910 server, or are building with only 1 PSU active so won’t see the bug.
The good news is that now its a reproducable bug, it’ll get fixed. Theres nothing worse than a bug that is specific to certain machines 🙂