Thursday, December 22, 2005

Veritas Checkpoints -

SNAPSHOTS are a point in time copy of a filesystem.

mount –F vxfs –o snapof=origfs[,snapsize=size] destination snap_mount_point

mount –F vxfs –o remount, snapof=origfs[,snapsize=size] destination snap_mount_point

umount snap_mount_point

Checkpoints are still better, they are persistent across a reboot and utilise fewer resources.

Primer by Par Botes

Here is a quick primer on checkpoints: (can you tell I really think
should be used instead of snapshots... :)

Checkpoints take
fewer resources than snapshots, they are persistent, you can
independently write to a checkpoint and they will survive a reboot.

Having said all of that... Both of them can handle multi terabyte
filesystem. Checkpoints will do it better and with fewer resource

Create checkpoint:
fsckptadm -n create "a_name" /from_running_filesystem

List checkpoints:
fsckptadm list /from_running_filesystem

Mount checkpoint (default is readonly):
(note that each checkpoint get a pseudo /dev/vx/dsk/dg/ name)
mount -F vxfs -o ckpt=a_name /dev/vx/dsk/diskgroup/volname:a_name
Mounting it read/write:
mount -F vxfs -o ckpt=a_name,rw /dev/vx/dsk/diskgroup/volname:a_name

Umounting the checkpoint:
Umount /checkpoint_mount

Page 68 and onwards in the vxfs administration guide is a really good
read on checkpoint functionality.


This was posted by Par on the veritas mailing list.


This is small place I'am trying to fill with computer literature, mostly about SAN and UNIX, well I'll also include Linux and a few notes on NAS too.

So if you are interested in the following, just keep peeking up this place for more

1. Veritas.
2. SUN Solaris.
3. HP Storage.
4. EMC(square).
5. Netapp Filers.
6. IBM AIX and IBM Storage.
7. Redhat Linux and other topics.
9. iSCSI
10. FCIP
.................................and more