On Thu, Jul 30, 2009 at 3:54 PM, Joseph L. Casale<jcas...@activenetwerx.com> wrote: > Anyone come up with a solution to manage the replication of ZFS snapshots? > The send/recv criteria gets tricky with all but the first unless you purge > the destination of snapshots, then force a full stream into it. > > I was hoping to script a daily update but I see that I would have to keep > track > of what's been done on both sides when using the -i|I syntax so it would not > be reliable in a hands off script. > > Would AVS be a possible solution in a mixed S10/Osol/SXCE environment? I > presume > that would make it fairly trivially but right now I am duplicating data from > an > s10 box to an osol snv118 box based on hardware/application needs forcing the > two platforms. > > Thanks for any ideas! > jlc > _______________________________________________ > zfs-discuss mailing list > zfs-discuss@opensolaris.org > http://mail.opensolaris.org/mailman/listinfo/zfs-discuss >
I came up with a somewhat custom script, using some pre-existing scripts I found about the land. http://www.brentrjones.com/?p=45 I schedule some file systems every 5 minutes, hour, and nightly depending on requirements. It has worked quite well for me, and proved to be quite useful in restoring as well (already had to use it). E-mails status reports, handles conflicts in a simple but effective way, and replication can be "reversed" by just starting to run it from the other system. I expanded on it by being able to handle A-B and B-A replication (mirror half of A to B, and half of B to A for paired redundancy). I'll post that version up in a few weeks when I clean it up a little. Credits go to Constantin Gonzalez for inspiration and source for parts of my script. http://blogs.sun.com/constantin/ -- Brent Jones br...@servuhome.net _______________________________________________ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
