Recovering a file system

File Service Administration Guide for Hitachi NAS Platform

Version

14.9.x

Audience

anonymous

Part Number

MK-92HNAS006-31

Following some system failures, a file system may require recovery before mounting. If required, such a recovery is performed automatically when you mount the file system. Performing recovery rolls the file system back to its last checkpoint and replays any data in NVRAM.

In extreme cases, when you mount a file system after a system failure, the automatic recovery procedures may not be sufficient to restore the file system to a mountable condition. In such a case, you must forcefully mount the file system, which discards the contents of NVRAM before mounting the file system.

Navigate to Home > Storage Management > File Systems.
Select a file system and click details to display the File System Details page.
If a file system displays Not Mounted in the Status column, click mount to try to mount the file system.
- If necessary, the automatic recovery processes will be invoked automatically. The file system was mounted successfully.
- If the automatic recovery fails, the file system will not mount, and the File Systems page will reappear, indicating that the file system was not mounted. Navigate to the File System Details page.
For the file system that failed to mount, click details to display the File System Details page. In the Settings/Status area of the page, the file system label will be displayed, along with the reason the file system failed to mount (if known), and suggested methods to recover the file system, including the link for the Forcefully mount option.
Depending on the configuration of your system, and the reason the file system failed to mount, you may have several recovery options:
- If the server is part of a cluster, you may be able to migrate the assigned EVS to another cluster node, then try to mount the file system. This can become necessary when another node in the cluster has the current available data in NVRAM that is necessary to replay write transactions to the file system following the last checkpoint. An EVS should be migrated to the cluster node that mirrors the failed node's NVRAM (for more information on NVRAM mirroring, refer to the System Access Guide. For more details on migrating EVSs, refer to the Server and Cluster Administration Guide.
- If the first recovery attempt fails, click the Forcefully mount link. This will execute a file system recovery without replaying the contents of NVRAM.
CAUTION:
Using the Forcefully mount option discards the contents of NVRAM, data which may have already been acknowledged to the client. Discarding the NVRAM contents means that all write operations in NVRAM (those write operations not yet committed to disk) are lost. The client will then have to resubmit the write request. Use the Forcefully mount option only upon the recommendation of customer support.