In most of circumstances, parity and media scan are background processes, user and storage admins won't need to take care of them. However, things could go wrong in your unexpected way. Here is my story.

Before number of media errors get to the threshold, controller will just mark the sector unusable. I had one drive had a lot 'stable but unusable' errors in the event log for long time, eventually it became 'impending drive failure', hot spare kicked in but recovery guru shows a lot 'unreadable sector' errors. In the meantime, the LUN can't be accessed from host.

Here is the solution to fix the problem, you can use it as a reference when you facing the similar problem. You may have all drives good, but parity goes wrong.

Note: Ask your vendor before you perform the following steps to your system.

Access Script Editor of Storage Manager

sudo - SMclient

Go to the Enterprise Management View, where Devices and Setup are located, you will then right click on storage > execute script,and in the box that pops up you can issue commands

Or, run SMcli command line

Replace the impending drive

Fail and replace the impending failure drive, see How to fail and replace impending drive failure, wait the reconstruction finish.

Parity Check

Issue the following commands in the script editor:

 check logicalDrive [dcs3700_lun0] parity
parityErrorFile="/home/smclient" priority=high verbose=true;

note: destination folder and file name can be changed    

Performing syntax check...
Syntax check complete.
Executing script...
check logical drive["dcs3700_lun0"] parity command has started.
Logging parity error at LBA: 19,968 to: /home/smclient/ParityError.txt.
Logging parity error at LBA: 19,973 to: /home/smclient/ParityError.txt.

Clear logical Drive Unreadable Sector errors

Issue the following command in the script editor again:

clear allLogicalDrives unreadableSectors;

Collect a new supportdata

The next steps are most likely going to be as follows, but please first collect support data

Repair Parity errors                                                             

Perform a parity repair on the affected volume by issuing the following command in the script editor. If no Parity Error file is generated, then no errors are identified. During this scan, if new errors are encountered, additional events might be logged against this drive.

 repair logicalDrive [dcs3700_lun0] parity
parityErrorFile="/home/smclient/ParityError.txt" verbose=true;

The command above will repair the parityerrors from the previous parity check

Verity errors

To verify that all errors have been resolved, perform a second parity scan on this volume. No further errors are encountered. Further, wait for the next background media scan to execute;

Try mounting the LUN

Try mount the LUN again, see if all problem fixed.


Comments powered by CComment