Well, this article sums up my three days of misery in an entirely objective way, L O L.
I should hasten to add that our (i.e. my) Oracle 10g upgrade was complete and available 2 hours before schedule. It was only then that we ran into, erm, memory problems on our production server, for which I am completely blame free. I was in from 00.00 – 18.30 on Sunday and 09.00 – 22.30 on Monday. And a Tuesday full of agony.
Normal service has resumed, Stu’s stress levels reduced significantly. Still, the article is pretty hilarious. 99% success?!
01-Aug-07 at 3:47 pm | Permalink
I sympathise.
> “We had huge confidence that the upgrade would go smoothly, so this really took us by surprise.”
That’s pretty surprising honesty, as it translates to “we had no fucking backup plan whatsoever”.
TOTALLY HOPELESS.
01-Aug-07 at 3:57 pm | Permalink
that is quite amusing
01-Aug-07 at 6:04 pm | Permalink
> “we had no fucking backup plan whatsoever”
Not strictly true.
We had 3 backout options, but they allowed the bank to open on the Sunday night therefore allowing online transactional activity all through the night (including a batch run) and once that happens backout is no longer a realistic option, we knuckle down and fix forward.
Everything OK now though thankfully.