#177956 - 2007-07-14 06:52 PM
Re: Translation, please
[Re: Lonkero]
|
Richard H.
Administrator
Registered: 2000-01-24
Posts: 4946
Loc: Leatherhead, Surrey, UK
|
It's 6 blades, each of which have two disks in a Raid 1 configuration.
On every blade, one of the disks is marked bad - if I reseat the disk it rebuilds and is good for a short while and then one of the pair is marked bad again, though not necessarily the same one that was originally bad 
The servers and disks are very new, and all disks failed at around the same time.
So the tech support guy is sending me 12 new disks to replace *all* the disks in the blades. This means I will have to swap out the currently failed drive, wait for the RAID to repair then swap out the other drive.
Sounds like a pointless exercise to me, but then as I've already been instructed to apply a firmware upgrade which isn't supported on these disks and watched as the program failed completely to do any useful update nothing surprises me.
|
|
Top
|
|
|
|
|
#177964 - 2007-07-15 02:35 PM
Re: Translation, please
[Re: Sealeopard]
|
Glenn Barnas
KiX Supporter
Registered: 2003-01-28
Posts: 4402
Loc: New Jersey
|
Last summer, I worked on a project where we used an HP blade chassis. When it arrived, the local admin unpacked it and assembled it. He fired it up, and it failed after running for a while. I dug through the trash and pulled out the documentation. Reading the quickstart guide, I found that if you install 4 of the six power supply / fan modules, you can't install 1,2,3,4 as he did, you need to install 1,3,4,6. I repositioned the power supplies, fan modules, and redistributed the blades, and the problem disappeared.
The point is - even minor misconfiguration in the blade chassis can cause problems across the blades themselves. It's possible that you received drives from a defective manufacturing lot, but my money would be on a bad or misconfigured cooling module, or a power supply that's impacting the +12v line in the chassis. Only the drives use 12v nowadays, and noise on that line can impact the rotational accuracy enough to trigger a drive "failure". With SMART drives, they report errors back to the controller that could be interpreted as a motor failure.
Glenn
_________________________
Actually I am a Rocket Scientist! 
|
|
Top
|
|
|
|
|
#177968 - 2007-07-15 11:39 PM
Re: Translation, please
[Re: Glenn Barnas]
|
NTDOC
Administrator
Registered: 2000-07-28
Posts: 11629
Loc: CA
|
And what happened to the days where an actual Engineer would come on site to determine the root cause and fix it on the spot. Ah, I suppose those days left when their profits went from 45% markup on everything to 3 or 4% markup on everything 
Well now that you have everyone's curiosity going Richard you'll have to keep us posted with the real cause and fix.
|
|
Top
|
|
|
|
Moderator: Arend_, Allen, Jochen, Radimus, Glenn Barnas, ShaneEP, Ruud van Velsen, Mart
|
0 registered
and 568 anonymous users online.
|
|
|
