MountV resource fails to online with error: Filesystem <drive_letter_or_mount_point> is not clean
Article ID: 100029334
Updated On:
Description
Error Message
VCS ERROR V-16-10051-9026 MountV:EV_SG-MountV-6:online:Filesystem at
* Log file = %vcs_home%\log\mountv_A.txt
Symptoms
- The MountV resource remains in an 'Onlining' status for a period of time, then eventually faults.
- The MountV resource remains in an 'Onlining' status for a period of time, then eventually faults.
- Many scenarios could result in this, so it's necessary to view the mountv_a.txt log file to determine the true cause.
Cause
- The file system has the NTFS dirty bit flag set. This is most likely caused by a force dismount or an unclean shutdown while volume was mounted.
Cause
- The file system has the NTFS dirty bit flag set. This is most likely caused by a force dismount or an unclean shutdown while volume was mounted.
Temporary Workaround
It may be necessary to bring the volumes/application online as quickly as possible. If so, the following temporary workaround can be used to all the cluster to bypass this issue. This will involve manually mounting the volumes from within the Veritas Enterprise Administrator (VEA) console.
1. Check the MountV resource properties and note the Volume Name and the Drive Letter or Drive Path.
2. Within VEA, locate the volume > right-click > File System > Change Drive Letter and Path...
3. Assign the proper Drive Letter or Mount Path that matches the MountV resource.
4. Wait up to 60 seconds for the MountV resource to probe in the Cluster Administrator console, or right-click on the MountV resource and choose the 'Probe' option. The resource should then show online.
At this point, the Service Group can be brought online successfully; however, if the MountV resources go offline, the same issue will occur when it attempts to online the resource the next time. In order to truly resolve this, follow the steps below.
Resolution
Method A
1. Check in the MountV_A.txt for the above error message to see which drive needs to be cleaned. (MountV_A.txt can be found in %vcs_home%\log folder)
2. Flush the service group on that node to stop the MountV resource from Onlining (right-click Service Group > flush)
3. From the problem MountV resource properties, set the AutoFSClean attribute to True.
4. Try to re-online the resource. It will run a Microsoft chkdsk /f on all volumes that are marked dirty and if successful, it will then online the resource.
Note: It is possible the resource could fault due to timeout as the chkdsk continues to run. In addition, the chkdsk operation is run in the background which makes it difficult to know the status. Due to these concerns, the following method provides steps to allow chkdsk to be run manually outside of the cluster.
Method B
1. Check in the MountV_A.txt for the above error message to see which volume(s) needs to be cleaned. (MountV_A.txt can be found in %vcs_home%\log)
2. Flush the Service Group on that node to stop the MountV resource from onlining (right-click Service Group > flush)
3. Save and close the configuration of the cluster config. This can be done from the Cluster Administrator console or by running the following command: haconf -dump -makero
4. From the command prompt, run: hastop -local -force. his will stop the cluster service without offlining any resources that are currently online in the cluster.)
5. Launch Veritas Enterprise Administrator and manually add drive letter to the problem volume (right-click volume > File System > Change Drive Letter and Drive Path...). And assign same Drive Letter or Drive Path as configured for the MountV resource.
6. From the command prompt, run the chkdsk manually on the problem volumes: chkdsk <drive_letter_or_or_mount_path> /F (e.g. chkdsk G: /F)
7. Once chkdsk completes on all problem volumes, go back into the VEA console and remove the manually assigned drive letters and/or drive paths. This will allow the cluster to bring them online so they remain under cluster control.
8. Restart the cluster by running the following command: hastart
9. Log into the Cluster Administrator console and bring the Service Group(s) online. The MountV resource(s) should now online properly.
Issue/Introduction
MountV resource fails to online in Storage Foundation for Windows HA cluster.
Was this article helpful?
Yes
No
Powered by
