Received work ticket. It was straight-forward: replace systemboard in server, bring it back online.
Okay, cool, I can do that! Done several of these!
Check out new board from spares cage
Go out to server location. Power it down. Remove it from rack. Chuck onto anti-static cart.
Pull out old systemboard, insert the new.
Return to rack, power up: green for half-second, amber for two seconds, green for half-second, repeat
Hrm.
Check out new board from spares cage
Replace first ‘replacement’ board with new.
Place server in rack: VOILA, online.
Set all required settings, Remote Management and what-not. Boot to Windows. Ping out and make sure everything worked okay.
Left server and returned to office. Remote Desktop into server, still works.
Finish ticket diary with ‘Completion’ notice…suddenly notice that RDC connection failed, unable to re-establish or ping.
GRR
Run out to colo: oh, hey, logged in via cached domain credentials. DIE DIE DIE.
Retrieve third motherboard. Re-set all settings again. Remarkably, this one works. At least I had another day to get the work done..
Damn you HP, be better at ‘refurbishing’ your systemboards before sending them back out to spare pools! …Or maybe can’t blame them for primarily replying on Foxconn parts. And Micron.
Nah. Cleansing flame.