Unsolved
This post is more than 5 years old
45 Posts
0
14647
April 13th, 2007 01:00
RAID 5 - drives failed on a PowerEdge 700 Help!
A problem booting up a poweredge 700 tonight.
It has 3 40GB SATA drives in a raid 5 array using the dell CERC SATA1.5/6ch card.
People were sayin they couldn't access shared files over the network (it's SBS 2003). I was coming over nd by the time I got there, someone had tried rebooting the server with the power switch. at this point there's a steady tone coming from the server. Watching the bootup, it shows the 3 drives in the array, the last column says smart errors and each drive has an N on its line.
But then the line after that says array - raid 5 FAILED and the system doesn't boot
going into the controller bios, it showed the array, and then pressing enter on that array, lists the 3 drives, but 2 are grey (drives 0 and 2) and drive 1 is white. I ran verify hard drive on each of the 3 drives and it came back veryfy completed for each one (no message of pass or fail - no news is good news?!).
again, this is a raid 5 and the raid status in the bios control screens is failed (not degraded).
In speaking with Dell, they had me take 1 drive out at a time with server off and reboot. e ach time, it says the drive is missing. In that process, the bios says there's a change to the configuation and he said to accept the change... it still shows the 3 drives in that array, but after reconnecting all drives, they all show as grey.
the dell rep said to press control R (enable / restore raid). it warns taht all data may be lost. I haven't done this yet... I am hoping someone here can tell me yess or no to that.
any ideas? 2 drives failing at the same time? but then all three drives pass the verfy test
It has 3 40GB SATA drives in a raid 5 array using the dell CERC SATA1.5/6ch card.
People were sayin they couldn't access shared files over the network (it's SBS 2003). I was coming over nd by the time I got there, someone had tried rebooting the server with the power switch. at this point there's a steady tone coming from the server. Watching the bootup, it shows the 3 drives in the array, the last column says smart errors and each drive has an N on its line.
But then the line after that says array - raid 5 FAILED and the system doesn't boot
going into the controller bios, it showed the array, and then pressing enter on that array, lists the 3 drives, but 2 are grey (drives 0 and 2) and drive 1 is white. I ran verify hard drive on each of the 3 drives and it came back veryfy completed for each one (no message of pass or fail - no news is good news?!).
again, this is a raid 5 and the raid status in the bios control screens is failed (not degraded).
In speaking with Dell, they had me take 1 drive out at a time with server off and reboot. e ach time, it says the drive is missing. In that process, the bios says there's a change to the configuation and he said to accept the change... it still shows the 3 drives in that array, but after reconnecting all drives, they all show as grey.
the dell rep said to press control R (enable / restore raid). it warns taht all data may be lost. I haven't done this yet... I am hoping someone here can tell me yess or no to that.
any ideas? 2 drives failing at the same time? but then all three drives pass the verfy test
I could update bios firmware (I have 1.02 and it's up to1.06) and the raid controller has an update to the firmware also. but I want to hear from experts here first!
No Events found!
babaganoosh
45 Posts
0
April 13th, 2007 01:00
at5147
884 Posts
0
April 13th, 2007 02:00
Performing the crtl+r will force all drive back into the array. Generally that is not recommend unless diagnostics have been run first or if you know that a drive drive before another. In this case, it can be assumed that the drives themselves are working properly and, though cannot verify if the drives did indeed go off at the same time, with the server becoming unresponsive as it did, it is likely they did.
There are some guesses that are being made here of course. The only reason that I can see you would want to have a drive removed from the system when performing the ctrl+r is if you knew that one drive fell offline before another. From my understanding, in the Adaptec controllers, when doing that all drives should be present unless there is reason to assume a drive failed diagnostics or one with offline before another.
The reason that you would want to have a drive removed in the latter of the two is the parity information of the array would be different. The controller would not know how to calculate it correctly between the drives, thereby corrupting the parity, or your data.
Just some thoughts that I have.
jeroenov
3 Posts
0
May 27th, 2008 08:00