On Mon, Jun 27, 2011 at 12:05 AM, Amit Tewari <at at amit-tewari.name> wrote: > ** > Hi there, > Snaphots were part of GlusterFS roadmap. > http://www.gluster.com/community/documentation/index.php/GlusterFS_Roadmap > > - *writable-snapshot* - *distributed writable snapshot translator* > > do we have some read-only snapshot mechanism in place/ in development/ in > development plan? > > Regards > Amit > Hi Amit, Sorry for the confusion. Thats our old roadmap page. John Mark (Director of Communities) will soon publish an updated roadmap. In the mean time, you can ask him such questions directly via IRC. Dave Garnett (VP Product Management) is also on the IRC most of the time. We wrote a file level snapshotting translator for VM images and killed it. We are re-investigating our snapshot design with continuous data protection architecture. Snapshots are best implemented at the block layer. They do not scale for large amounts of data. Also in the cloud space, admins wanted to completely automate snapshots (point in time without worrying about running out of snapshot counts). Best implementation of snapshots is NetApp's WAFL. Even their snapshot can scale only up to 256 and 16TB volume capacity. Here is what we are hoping to achieve: * Unlimited snapshots on a volume or per file basis. Think of it as versioining filesystem instead of block layer snapshots. * Scalable to multiple petabytes. * No admin intervention. Set once and forget it. For example: snapshot every 30 mins and merge one full copy every 16 days. * Giving users ability to view/restore versions of their files and folders without requiring admin's help. It doesn't matter if we call this snapshot or CDP or continuous backup. It brings continuous versioning like capability and meets the requirements :) -- Anand Babu Periasamy Blog [http://www.unlocksmith.org] Imagination is more important than knowledge --Albert Einstein -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://gluster.org/pipermail/gluster-users/attachments/20110627/fe4e9037/attachment-0001.htm>
