CHKDSK

If you must run CHKDSK on a volume that is being replicated by SIOS DataKeeper, it is recommended that you PAUSE the mirror before initiating the CHKDSK. After running CHKDSK, CONTINUE the mirror. A partial resync occurs (updating those writes generated by the CHKDSK) and replication will continue.

Note: The bitmap file (for non-shared volumes) is located on the C drive which is defined by BitmapBaseDir as the default location. Running CHKDSK on the C drive of the Source system will cause an error due to the active bitmap file. Therefore, a switchover must be performed so that this Source becomes Target and the bitmap file becomes inactive. The CHKDSK can then be executed on this system as the new target (original source).

How to run a CHKDSK in a Highly Available Manner

If your disks are experiencing corrupt files, crashes via improper shutdowns, including the following Windows Event Viewer IDs, consider the usage of the Windows utility CHKDSK.

• Event ID: Bad block on a disk
• Event ID: 55 File System Corruption to include suggesting running a CHKDSK
• Event ID: 98 A volume is required to be taken offline to run a CHKDSK
• Event ID: 153 Disk error reflecting where an I/O operation was retried

These steps should be performed on Server1 and Server2 (Source/Target).

DataKeeper Administration

On Server2 (Target/Passive node), launch the DataKeeper GUI (datakeeper.msc).

• Select Pause and Unlock Mirror.

Windows Server Administration (Storage)

Launch an elevated Administrator command prompt.

Without immediately performing a FULL CHKDSK, verify the status of the offending volume(s) by running the following command:

chkdsk s: (or the drive letter at fault)

The summary below reflects no errors:

Windows has scanned the file system and found no problems. No further action is required.

5224447 KB total disk space.
    628896 KB in 138 files.
       112 KB in 25 indexes.
         0 KB in bad sectors.
     10959 KB in use by the system.
     10128 KB occupied by the log file.
   4584480 KB available on disk.

1306111 total allocation units on disk.
   1146120 allocation units available on disk.
Total duration: 54.91 milliseconds (54 ms)

• If you’re receiving bad indexes, bad sectors or other errors, execute a FULL CHKDSK. The syntax will be as follows for Volume S:

chkdsk /F S:

*This is being performed on Server2 (Target/Passive node), the duration of the scan can vary from several seconds to several hours based on the severity of disk(s).

DataKeeper Administration

Once the chkdsk has completed, providing the desired result, launch the DataKeeper GUI (datakeeper.msc).

Select Continue and Lock Mirror.

Failover Cluster Administration (cluadmin.msc)

If the desired result has been achieved, you can now perform a Move/Switchover, (e.g. of your Role and associated resources from Server1 to Server2).

Server2 now becomes the Source/Active node and Server1, the Target/Passive node.

Repeat the aforementioned steps on Server1.

*If you’re not able to fix the corrupt disk via CHKDSK, please consider replacing, also in a “Highly Available” manner.