Hi,
On Mon, Aug 16, 2010 at 05:19:24PM +0000, Andy Smith wrote:
Hi,
On Fri, Aug 13, 2010 at 02:16:52PM +0000, Andy Smith wrote:
The hardware vendor is convinced it is still a
RAM mismatch issue. I
am skeptical given that the RAM used so far was matched to the
motherboard and of the same type, but they are insistent and have
loaned me some that they're sure of, to test.
Server locked up again at approximately 1704Z (18:04 UK time),
proving not an issue with the RAM. I need to discuss with the vendor
but there will need to be further work, very soon. I will update you
of when this is to take place.
I've run out of patience with the vendor on this (latest hoop: "run
memtest").
On Monday 23rd I'm installing more servers and provided all goes
well I intend to put this issue to rest as far as customers are
concerned by returning on Tuesday 24th to swap the disks from
urquell into one of the new servers.
It will from then on be using entirely new hardware aside from the
disks, leaving me plenty of times to do all the pointless things the
vendor wants me to do without further affecting your services.
Lesson learned; buy more hardware, not hardware support.
Should there be further issues before Tuesday, I will have to power
cycle it; any other action is going to involve more downtime than a
reboot.
I am planning to carry out the work at 0630Z (7.30am UK time)
Tuesday 24th August and I expect it to take less than 30 minutes,
with no action necessary by yourselves. It will involve shutting
down your VPSes, shutting down the server, taking the disks out and
placing them in the other server, booting up the other server and
then your VPSes will also start again.
Cheers,
Andy
--
http://bitfolk.com/ -- No-nonsense VPS hosting