You’ve seen the limbo line, with each time around going a little lower, until the player crashes. Well, the past two weeks I’ve been playing “IT Limbo”.
If you’re an IT person, and responsible for a lot of gear, you’ll know about this. It all begins when an important system fails, and you have no choice to hammer on it till it comes up.
We ended up in a situation where we had to move a bunch of bladeservers and EMC VNX arrays. Such devices do not like to be moved, and require lots of careful planning. We did our due dilligence, had a carefully devised plan, and on the appointed day, wespent many hours dropping volumes, luns, replication and mirroring links, failing things over to a secondary, etc. We run in a system where every volume has a replication partner, and a third data storehouse for dead data (backups). Trouble is, when moving, one of the hot partners has to be shut off, leaving you with one hot array.
We had successfully finished the job, were cleaning up, and preparing to leave for the day, when 100’s of panicked calls started coming in - “the XYZ is down! OMG!” We ran into the data center, where the movers were just leaving, and began frantically trying to figure out what happened. Didn’t take long - one of the movers had accidentally unplugged 8 twinax 10gig cables, and dropped our remaining array (1 petabyte worth) by disconnecting it from the world. So we started plugging the cables back in.
And no ESX server would connect. After going through basic troubleshooting, nothing was working, and we started dialing support and lining up troops.
We call it going around the world. When you get on a support call, and as each tech finishes his/her shift, they transfer you to the next open support center, usually somewhere to the west of you. This time, we went around the world with EMC for five days. 24/7.
We set up cots in our offices. I managed 9 hours sleep in 5 days. Everyone who saw me thought I had cancer, or some dread malady. I finally passed out one morning, and the staff was unable to wake me. I slept 7 hours. So evidently, 9 hours in 5 days was as low as I could go. Limbo!
BTW - system is back online, and all is much better now. I’m just about desperate to do my things now, or to have any time away from the office. Can’t wait to get back to my projects.