This post is more than 5 years old
5 Posts
0
59186
Issue with Perc 5/e and MD1000 Enclosure
We have a dell Poweredge 1950 running Windows Server 2003 SP2. The server has a Perc 5/e card that is connected to an MD1000. The MD1000 has 6 700GB SATA drives in a raid 5 and 6 2TB SATA drives in a raid 5. It has been working fine for years. We use the drives for backups. This morning I noticed all of our backups had failed last night. When I went to look I noticed that the drive letters were missing from windows. I checked the windows event log and noticed that there was an error .
SCSI sense data Sense key: 3 Sense code: 11 Sense qualifier: 0: Physical Disk 0:0:10 Controller 1, Connector 0
About 2 hours later there was a percsas event:
The driver has detected a device with old or out-of-date firmware. The device will not be used.
Then there are some errros about the perc 5/e having disappeared from the system without being removed properly.
Long story short is that I can't access the data. In dell open manager it shows that one of the raid 5 arrays are degraded but all the physical drives are still there and appear healthy. I can also see the perc 5/e card and that the md1000 is connected to it.
I haven't tried anything because of fear of losing data. Should I reboot the server and see if the drives appear again or should I try something else first to ensure that my data is safe? This has been running as is for 4 years and we haven't updated anything that would have caused a firmware error.
Dev Mgr
9.3K Posts
1
August 16th, 2012 07:00
First; I assume you have backups of this data (on tape or so), just in case.
After this assumption, I'd suggest indeed to do a normal shutdown of the server. Then power down the MD1000 (both power supplies). Give the MD 30-60 seconds to spin down the drives and fans. Power up the MD1000 again, give it 2-3 minutes to spin up the drives (ignore any fault lights for now). Then power up the server and see if you can see the drives (either in the PERC bios or in the OS/OpenManage). If not, shut down again and move the SAS cable to the other controller on the MD1000 to see if it's maybe just the EMM (be sure to connect to the in-port, not the out-port).
DELL-Chris H
Moderator
Moderator
•
8.4K Posts
0
August 16th, 2012 07:00
Alfonsovilar,
If not the EMM as Dev suggested it may be due to you not having updated the controller in 4 years. What you will need to do is update and then see if the issue resolves itself. Now since it hasn't been updated in so long you don't want to jump at the most recent firmware update, we need to step up the firmware to current. Here is the Perc 5/e driver and firmware updates.
Perc 5/e driver -
<ADMIN NOTE: Broken link has been removed from this post by Dell>
Perc 5/e firmware - www.dell.com/.../poweredge-1850
Go to OTHER VERSIONS on the page and then find your current version, you can then start running the updates up to the most recent. You can skip and run every other update.
Also the updates I provided need to run in the order listed, as well as they are designed to be run from within the OS.
Let me know how it goes.
alfonsovilar
5 Posts
0
August 16th, 2012 08:00
Thanks for the quick replys. I was taking a quick peek at the troubleshooter for the md100 and it says the first thing I should try is rebooting the server without powering down the md1000. Is there any risk to trying that as a first step?
alfonsovilar
5 Posts
0
August 23rd, 2012 04:00
So I shut everything down as suggested and powered everything back up. The windows drive letters came back online. Disk 10 tells me that it's online but that the configuration is foreign. Therefore one of the arrays appears as degraded. Is there a way I can clear the foreign tag on it so that it can be added to the array again?
Also with regards to the firmware upgrades for the perc 5/e card from within the OS... Do each of these incremental upgrades require a reboot? Because of the degraded status of the array right now I can't reboot the server remotely. I'd rather do these at night if possible without having to travel onsite.