This post is more than 5 years old
3 Posts
0
10412
PE2950 PERC5/i RAID-5 question
We have a 3 disc RAID5 array and disc 0:0:1 is showing "corrected" media errors...
These information items show up when a patrol read runs:
"2271
2266
2095 "
Used this link here to check the event ids....
The reccomended action is none for all... also I noticed that when the patrol read was running disc 0:0:0 I think it is does not show any activity the TOP LED is green but I only ever see the other two discs showing activity, granted I don't stare at it and, when I check open manage the disc's are all green and the virtual disc's are green as well.
Back around 12/22/2010 I swapped 0:0:0 as there was an event saying there was a prediticted failure. the array rebuilt w/out issue and it has been fine...
Should I worry? or is this normal? it is an exchange 2003 server so I really don't have to loose the array.. anyone experenced this?
TIA,
-c
DELL-Daniel R
57 Posts
1
February 23rd, 2011 09:00
Media errors being corrected during a patrol read is normal, and is what the patrol read is designed to do so problems like the predicted failure drive don't happen. The fact that the errors are being corrected is a good thing. If you have further concern about the hard drives in the array a quck diagnostic test can be run on the drives through Windows.
The link to the Online Diagnostics is: http://ftp.us.dell.com/diags/dell-onlinediags-win32-2.15.0.169.exe Once installed, please go to Programs> Online Diagnostics> GUI, wait for the devices to discover. Once all devices are discovered, select only the hard drives on the left hand side. The hard drives may be listed under SCSI Enclosures or PERC Controller. Once the hard drives are selected, select Quick Test on the bottom right hand side of the GUI Window, and then click on Run Test.
Also as a general rule I suggest making sure the driver and frirmware are up to date on the RAID controller as well as the firmware for the hard drives.
pcmeiners
4 Operator
4 Operator
•
1.8K Posts
0
February 23rd, 2011 09:00
"disk 0:0:1 is showing "corrected" media errors...
sector errors were found and remapped to "spare" sectors. If future Patrol Reads runs do not find/remap more errors your fine for now. If it continues to find more errors eventually the drive will be marked "bad" , depends how many sectors it remaps over time. Not unusual to get these errors as drives age. If every PR run find errors, expect to replace the drive in the near future. If you see two drives on a PR run produce errors, then more errors on proceeding runs, time to get cautious. I believe in having a hot spare drive in place for critical servers, so if you have the resources I would go for that.
riggsdlshs
3 Posts
0
February 23rd, 2011 14:00
So far it is only the one drive...
Can you even add a spare drive to an active array without disrupting the server operation or harming the array?
ok, so my last question... what about the 0:0:0 bottom LED not blinking? if open mangage says everything is ok should I not worry about that? I am just afraid that if 0:0:1 goes that my array will fail faster than a bank. Is this a valid concern?
thanks again!
-c
pcmeiners
4 Operator
4 Operator
•
1.8K Posts
0
February 23rd, 2011 16:00
Yes, place drive in slot with the machine powered up, open OMSA ,then follow the instructions in the link, ...
http://support.dell.com/support/edocs/software/svradmin/1.9/en/stormgmt/vdtkdhs.html#188087
If you want to be particularly careful about the hotspare being same firmware as the other drives, download the particular firmware, update it , then add it as a hot spare.
0:0:0 bottom LED not blinking" Normal
Your OK, did not mean to scare you, just monitor the drives... remember you can lose one drive, the array will run in degraded mode (slows a bit) until the bad drive is replaced and the array rebuilds in the background, if you don't have a hot spare, then replace the "bad" drive post haste,. If a drive go bad, unless the alarm is disabled, you will know. Basically after you disable the alarm, you want to replace the drive as fast as possible, not weeks or months later as I seen done. With a hot spare it is automatic.
riggsdlshs
3 Posts
0
February 23rd, 2011 16:00
Excellent, that is good info to have - I wish I would have thought of that sooner for the extra drive.
and yes I watch it like a hawk check it on and off up until I go to bed... I just wanted to make sure the array was indeed ok. As we are in the process of moving off that server so I just need to live a little longer.
Many thank yous!
-c