+1 On Wed, Jul 1, 2015 at 4:53 PM, Nick Dimiduk <[email protected]> wrote:
> This strikes me as a reasonable (and, err, surprising-that-it's-necissary) > request. +1 > > On Wed, Jul 1, 2015 at 3:46 PM, Andrew Purtell <[email protected]> > wrote: > > > +1 > > > > > > On Wed, Jul 1, 2015 at 3:27 PM, Stack <[email protected]> wrote: > > > > > +1 on making request. > > > St.Ack > > > > > > On Wed, Jul 1, 2015 at 2:33 PM, Sean Busbey <[email protected]> > wrote: > > > > > > > On one of our open issues about Hadoop versions, one of the Hadoop > PMC > > > > members mentioned that the 2.6.z line wasn't planning any additional > > > > releases[1]. > > > > > > > > I'd like us to request, as a downstream community, that the Hadoop > > > project > > > > plan for maintenance releases on this line given the non-production > > > status > > > > of 2.7.0, unevaluated quality of further 2.7 releases and the unknown > > > > status of a 2.8 release. > > > > > > > > Right now, there's substantial evidence from our Elliot that we > should > > be > > > > pushing our users from the 2.4/2.5 releases onto 2.6. At the moment, > > > 2.6.0 > > > > contains a couple of critical bugs that effectively prevent the use > of > > > HDFS > > > > transparent encryption[2]. Now, that feature isn't needed but it's > nice > > > to > > > > have as an operational alternative to our own implementation. And the > > > > current bug _destroys_ HBase clusters, so the consequences for the > > > curious > > > > are severe. > > > > > > > > That specific issue aside, however, as a system that runs on top of > > > Hadoop > > > > we impose on our downstream users a dependency on that project. > Regular > > > > maintenance releases are critical to easing long term operational > pain, > > > so > > > > we should proactively look out for them by prodding our less stable > > > > upstream dependencies. > > > > > > > > > > > > [1]: http://s.apache.org/MTY > > > > [2]: HADOOP-11674 and HADOOP-11710 > > > > > > > > -- > > > > Sean > > > > > > > > > > > > > > > -- > > Best regards, > > > > - Andy > > > > Problems worthy of attack prove their worth by hitting back. - Piet Hein > > (via Tom White) > > >
