at work: how one can screw up heartbeat cluster?
in many ways… first by mistyping configuration so both nodes think they are preferred masters. then.. by cross-connecting power cables from remote power control to nodes of the cluster in the way that slave kills itself instead of killing master. grate… what else? at least it was not my fault… by installing 3ware monitoring software for raid arrays… s
oftware that tends to hang whole machine in some very special configuration [ 3 degraded raid1 arrays.. ]. and all of this to achieve higher availability… i want to believe..
all of this was happening during last Friday afternoon that become Saturday’s early morning spent in a datacenter so
mewhere out of the town. and it was a lot of fun besides… taking walk home at 2.oo am for ~10km was not that smart after all. but it was not enough for me. i’ve just bought a proper [hopefully] bike and bad
ly needed to get a test drive. so i took it… 20 km around the city in -8 degrees. not smart… at all. so both should explain why i feel sick for last few days.
at home i’ve finished customizing small thigny for techgsm – bulk orders based on quite nice dhtml.
and besides all of those company has just supplied me a housemate for next 8 months. so far so good, we use IM on daily basis to communicate. not much domestic violence so far ;-]