Now that both Hadoop 1.0 and ZK 3.4.2 have been released I think we should ship with them.
Apart from that I'm +1. I even let a LoadTester job run on my 15 nodes during the holidays (287 hours now) and it's still crushing my cluster as much as it did in the beginning without any failure. J-D On Thu, Dec 22, 2011 at 4:22 PM, Stack <[email protected]> wrote: > The third hbase 0.92.0 release candidate is available for download: > > http://people.apache.org/~stack/hbase-0.92.0-candidate-2/ > > I've posted a secure and an insecure tarball built by jenkins. > > This RC bundles hadoop 1.0.0rc3 and zookeeper 3.4.2rc0. > > As said previous, HBase 0.92.0 includes a raft of new features > including: coprocessors, security, a new (self-migrating) file format, > distributed log splitting, etc. There's been > 600 fixes since 0.90. > See the list here: http://su.pr/1VZzl5 > > See the hbase manual for the low-down on what hadoop version this > release will run on: http://hbase.apache.org/book.html#hadoop > > There is no migration necessary. A shutdown and restart after putting > in place the new HBase should be all thats involved (A rolling restart > from 0.90.x to 0.92.0 will not work). That said, once you have moved > to 0.92.x, there is no going back to 0.90.x after the transition has > been made. > > Should we release this candidate as hbase 0.92.0? Take it for a spin. > Check out the doc. Vote +1/-1 by January 3rd, 2012. > > Happy Holidays. > Yours, > The HBasistas > > P.S. If you are thinking of going to 0.92 anytime soon, help us out w/ > the testing if you can. This period while we are putting up release > candidates is the best time for trying stuff out because if you find > an ugly bug, it'll be fixed pretty tout de suite
