Ethan Erchinger wrote: >> While I'll give you props for coming up with something innovative I'd >> strongly recommend against this type of solution. Manipulation like >> this really is flirting with a panic. Adding and removing devices >> too frequently on FC is in my experience dangerous in production >> environments. > Solaris (from my understanding) is probably the best system in the > world for handling FC environment, so this is disappointing news.
Solaris is the best system in the world, but you've seen it, I've seen it, sh*t happens, and Solaris is unforgiving when it comes to potential data corruption. I have confidence in the code, but I'll admit that when I quiescence a FC link my heart skips a beat, and it multiplies with the number of LUNs. :) >> >> My philosophy with FC on Solaris in a production environment is to >> set it and step away. If you can utilize ethernet, particularly with >> ZFS send/recv I think you'll be happier and avoid the cost of ibbackup. > We've thought of using snapshots and send/recv, but have backed away > from it for a few reasons. A) It doesn't perform any real validation > on the data as mentioned in a previous email. B) As stated all over > the place, this should not be used in place of a proper backup > environment. BUT, it's probably more reliable then what I'm trying to > do now. I'll disagree with the idea that send/recv shouldn't be used in place of a proper backup environment. One thing that really bugs me is that no conventional backup solutions have yet included ZFS send/recv capabilities, but its a radically different approach that what they've done in the past. Lets face it, if you dig into the guts of NetBackup whats at its heart is GNU tar. The most popular non-commercial solution is rsync. ;) For instance, the legendary Theo /Schlossnagle/ created Zetaback (https://labs.omniti.com/trac/zetaback) a really kool backup solution utilizing ZFS send/recv. No doubt you're aware of Mr. Jason Williams SnapBack (http://blogs.digitar.com/jjww/?itemid=56). And at Joyent I've been working on "RedSnapper" which uses ZFS send/recv for distributed backups in the cloud. When it comes to backup we all get into false assumptions about how things are "supposed to work", but at the end of the day your just making a copy of data in the easiest, least impactful, and fastest way possible. Simplicity is key, imho, and "simplicity" and "LUN re-map" don't jive. ;) Of course, as your well aware, the only trick involved with InnoDB datafiles backed up online via ZFS send/recv is that your gonna have to do crash recovery if you go back to them or use them elsewhere, hense where SnapBack just rules. Anyway, long winded way of saying don't give up on ZFS send/recv easily. :) Good luck man! benr. _______________________________________________ storage-discuss mailing list [email protected] http://mail.opensolaris.org/mailman/listinfo/storage-discuss
