Data integrity check and recovery using the vpfsd service

The vpfsd service performs a data integrity check in the background according to the data changes. An event is sent to the NetBackup primary when file corruption is detected, which includes the metadata and the data corruption (automatic process).

The corrupted file is moved to the following location:

/mnt/vpfs_share/<share dir>/<share ID>/.vpfs_corrupted

A syslog and a history log are created for the file. The backup of the universal share skips the files that have an issue.

Apart from the automatic process, you can use the vpfsck command to manually verify and detect the file corruption, and the corrupted file is moved to the following location:

/mnt/vpfs_share/<share dir>/<share ID>/.vpfs_corrupted

Only local VPFS is supported.

For example:

--verify_data  --share_id <share id> 
 [--t <file creation time older than (int) hours>]
 --check_so    --share_id <share id> 
 [--so <so fp, find files that contain this so>] 
 [--file_path <check single file or folder, 
 path relative to /mnt/vpfs/<share_id<0:4>>/<share_id>>]
 --verbose  optional

All the options after --share_id are optional.

The process is logged at <storage>/log/vpfs/vpfsd/vpfsck.log and the option --verbose enables the debug log for vpfsck.