24.2.1 Backing up using the sfcdump(1M)

PRIMECLUSTER Global File Services Configuration and Administration Guide 4.2 (Solaris(TM) Operating Environment)

Part 3 Global File Services Shared File System

> Chapter 24 File System Backing-up and Restoring

> 24.2 Backing Up and Restoring using sfcdump(1M) and sfcrestore(1M)

24.2.1 Backing up using the sfcdump(1M)

This section contains an example of backing up an entire GFS Shared File System specified in the /etc/vfstab file shown below. Here, /dev/rmt/0 is the backup destination tape drive.

% cat /etc/vfstab
#device                   device                     mount  FS   fsck  mount   mount
#to mount                 to fsck                    point  type pass  at boot options
...
/dev/sfdsk/gfs/rdsk/volume1 /dev/sfdsk/gfs/rdsk/volume1 /mnt   sfcfs  -   no      -
...

The character special file in the representative partition of the GFS Shared File System is /dev/rdsk/c0t10d0s0. The mount point is /mnt.

1. Set for super user.

% su

2. Unmount the GFS Shared File System to be backed up, so that the GFS Shared File System is not updated.

Even if the GFS Shared File System is mounted, a backup can be performed. The backup must be performed while the GFS Shared File System is not being updated. If the file system is updated during the backup, the backup contents may be invalid.

Note that sfcdump fails with the following message output if unmounted of the GFS Shared File System is activated and the mounted of the node that performs backup is stopped:

DUMP: Warning - /dev/FJSVgdhd/rdsk/gdhd0s0 is mounted on other node(s) or requires a file system check. Stop the file system or run fsck.

DUMP: The ENTIRE dump is aborted.

In such cases, unmount the GFS Shared File System or perform the backup after mounting the file system on the node performing the backup. If the above message is output when the file system is unmounted, the file system integrity may be lost. Use fsck to restore the file system.

For details on repairing a file system, see Section 23.2, "Checking a File System for Consistency and Repairing It."

3. Define a tape drive to backup a GFS Shared File System.

# sfcdump 0ucf /dev/rmt/0 /dev/sfdsk/gfs/rdsk/volume1
DUMP: Writing 63 Kilobyte records
DUMP: Date of this level 0 dump: Tue May 23 11:20:14 2000
DUMP: Date of last level 0 dump: the epoch
DUMP: Dumping /dev/sfdsk/gfs/rdsk/volume1 (kaede:/mnt) to /dev/rmt/0.
DUMP: Mapping (Pass I) [regular files]
DUMP: Mapping (Pass II) [directories]
DUMP: Estimated 282 blocks (141KB).
DUMP: Dumping (Pass III) [directories]
DUMP: Dumping (Pass IV) [regular files]
DUMP: 250 blocks (125KB) on 1 volume at 1838 KB/sec
DUMP: DUMP IS DONE
DUMP: Level 0 dump on Tue May 23 11:20:14 2000

Each character of 0ucf in the first parameter has the following meaning:

0
This is the backup level. The number 0 means a full backup. A number between 1 and 9 means an incremental backup. In this example, the file system is backed up as an incremental backup from the latest backup at a lower level than the specified backup level.
u
The dump record file, /etc/sfcfs_dumpdates, is updated. Specify this character for full backup or incremental backup.
c
The contents are backed up onto tape.
f
The write destination is specified. /dev/rmt/0 is specified as the write destination.

Contents Index