This is the home page of the Cluster Snapshot Block
Device. This is a virtual block device that runs
under Linux's device mapper subsystem, to allow you to create up to 64
read/write snapshots of a cluster filesystem (e.g., GFS) mounted on a
shared block device. Any or all these snapshots can be mounted
by themselves or simultaneously with the original volume.
What is this for? Why, backup of course. A
snapshot block device allows you to back up a snapshot of your
filesystem, while you continue to use your filesystem. A cluster
snapshot block device lets you do that with a cluster filesystem, as
well as a normal filesystem.
is the mailing list for cluster-related questions and discussion.
Whenever the development source code repository is updated,
email is sent to the cluster-cvs mailing list. This list is run using ezmlm.
Project TeamThat would be me!
StatusThe cluster snapshot is now feature-complete and lightly tested. Some subtle aspects of failover still remain to be implemented. The userspace server is coded in a very non-parallel way, and as a result is prone to bottlenecks on certain loads. The interface protocol is still subject to change. The whole thing does appear to work, and has been lightly tested, but not stressed.
A set of unit tests and regression tests is in preparation.
Here is the latest tarball, to be built against kernel 188.8.131.52. This code is suitable for unit testing and experimenting, not live use. Tracing output is on by default. Follow the directions in the README.
Note Note Note! This is still pre-alpha. Do not
use it on a filesystem you care about.
To get the latest source from CVS:
Just hit enter at the login password prompt.
Channel #linux-cluster on freenode