Heh, seems my blog post also spawned some internal conversations here too. After a review of the current architecture, we did some digging and determined that the main bottleneck (for the 3 days scenario) was in our NetBackup setup. After changing some "undocumented features" (http://seer.entsupport.symantec.com/docs/183702.htm), it looks like we'll be able to get everything to tape in around 8 hours. We are doing some further tests on straight ZFS send / receive between 2 solaris machines, which is the "24 hours" scenario in my post. We feel this should have been faster as well, especially now that we have backups to tape looking faster than 24 hours, but haven't put a finger on why. (Certainly one potential issue is that we're sending these snapshot from a very active production machine, so it's competing for resources). Oh, one additional note for those doing maths at home, the actual amount of data is closer 1.6TB.
Robert Treat Jignesh K. Shah wrote: > Hi Robert, > > Can you share the details where ZFS send takes three days? > > I hope to find some solution to help you with your ZFS snapshots and > which also in general will help PostgreSQL community for DR solutions > using ZFS. > > Thanks. > > Regards, > Jignesh > > > Nathan Kroenert wrote: >> Indeed - >> >> If it was 100Mb/s ethernet, 1TB would take near enough 24 hours just >> to push that much data... >> >> Would be great to see some details of the setup and where the >> bottleneck was. I'd be surprised if ZFS has anything to do with the >> transfer rate... >> >> But an interesting read anyways. :) >> >> Nathan. >> >> >> >> Nicolas Williams wrote: >>> On Wed, Apr 09, 2008 at 11:38:03PM -0400, Jignesh K. Shah wrote: >>>> Can zfs send utilize multiple-streams of data transmission (or some >>>> sort of multipleness)? >>>> >>>> Interesting read for background >>>> http://people.planetpostgresql.org/xzilla/index.php?/archives/338-guid.html >>>> >>>> >>>> >>>> Note: zfs send takes 3 days for 1TB to another system >>> >>> Huh? That article doesn't describe how they were moving the zfs send >>> stream, whether the limit was the network, ZFS or disk I/O. In fact, >>> it's bereft of numbers. It even says that the transfer time is not >>> actually three days but "upwards of 24 hours." >>> >>> Nico
