Virtual Disk bad blocks

Thorwaldhallvardsson,

With Open Manage Server Administrator installed you would need to run the following omconfig CLI command to clear the blocks.

omconfig storage vdisk action=clearvdbadblocks controller=id vdisk=id

Where id is the controller ID and virtual disk ID as reported by the omreport
command.

To obtain the values for controller ID and virtual disk ID, type 

omreport storage controller  to display the controller IDs,

and then type 

omreport storage vdisk controller=ID  to display the IDs for the virtual disks

attached to the controller.

Let me know if this helps clear the blocks.

Hi,

I have the same option from the gui. The trouble is it does say that it is going to clear bad blocks but not recover the data. So let's say if half of the virtual disk has bad blocks then I will loose 50% of the data. Hence my question was how to find out how many bad blocks does it have ? If it's 3 or 4 I won't be bothered but if there are millions then that is something to be worried about, isn't it ?

Thank you for your reply.

I experience exactly the same issue with my SBS 2011 server. Same setup and behaviour.
Did you manage to find a solution for this?

This is what happened to us. Upon replacing the faulty drive in the server and it doing the rebuild it found bad blocks. The server was still working but we were unable to do a full backup. We installed shadowprotect which has an option to backup and ignore bad sectors. We did this so at least we had a backup. We then went into the controller using the dell management studio and said mark bad sectors. The server was fine and we did a checkdisk and rebooted and everything appeared fine. We could also do a backup fine with windows backup and shadowprotect. We did find about a month later that we did have a corruption in one minor database file. We cant be one hundred precent sure that it was from the bad block but I believe it was. We were able to restore it from an earlier backup before the issue and everything now appears ok. I believe the issue of marking the bad sectors is only a problem if data exists on the sector. If your backup is failing there is a good chance that data does exists in the sector so make sure you have a backup from earlier in the period and keep it just in case you only find out later was is corrupted. Also be careful if the bad block exists in the file allocation table on the drive it may wipe the drive when you mark the bad blocks that is why you should try to get some sort of a backup before you mark the bad sectors. I read somewhere else where something like this happened. If you have any questions please post and I will try my best to answer 

clearing the bad blocks WILL cause data corruption. the severity of that corruption is going to be almost impossible to gauge based on the number of bad blocks. Further, the reason for bad blocks developing, whether it's bit-error growth, unstable heads, or read errors due to other causes, is important, because it's likely bad blocks will continue to develop. 

Using a utility that will skip bad blocks is relatively safe, but will still put stress on any unstable drives and risk more data corruption.

A better option would be to make images of all stable drives and then make images of any unstable drives as best as you can skipping bad blocks.

The safest possible option is to use a professional data recovery lab(if time allows) and get a standard free diagnostic done.

www.dell.com/.../EN Partners

Not sure what your trying to do, bad block are bad blocks, nothing will be written to them, or read from them. All the "clearing" or wishing will not make them better. In the process of being marked bad, there is an attempt to move readable data to a new location, once done or if the data is corrupt the block(s) are marked and can not be used. There is an attempt by the controller to test and recover the blocks before they a marked bad. If you were able to clear them as be marked bad, the controller will only remark them as unusable block, plus a good possibility of failing the array in the process.. What you need to be concerned with are blocks which are becoming difficult to read, but have not failed yet, For that you can run Consistency Checks or better still Patrol Reads,

How can I do this cleaning with the racadm command?

Hi, thanks for choosing Dell. Radcam is iDRAC CLI but you need OMSA CLI.
https://dell.to/3uVNcAC
To clear these bad blocks, execute the Clear Virtual Disk Bad Blocks task. This can be done in the OMSA GUI or use the cli command:
omconfig storage vdisk action=clearvdbadblocks controller=id vdisk=id