So, some update on that (sorry for long delay). I've posted my results, conclusions and thoughts in the comments to HBASE-16232, please check it out for details.
TL;DR based on findings and observations outlined there I believe we shall not consider this one as a release blocker and move with the release. I'm stabilizing and cleaning up the branch again for the release. Will send more updates as it goes. -Mikhail On Wed, Aug 3, 2016 at 11:34 PM, Mikhail Antonov <[email protected]> wrote: > Lately the only things blocking release have been integration test issues. > > First it was HBASE-16074 which took quit some time to address. > Now HBASE-16232 is the only thing blocking it > (ITBLL running on sufficiently large cluster with enough regions looses > data. For the record, I once saw the lost keys > on the build off branch-1.2 few weeks ago so not exactly sure yet when > the regression creeped in). > > -Mikhail > > On Thu, Aug 4, 2016 at 8:46 AM, Stack <[email protected]> wrote: > >> Thanks for asking Nick. >> >> Our Mikhail is on holiday and last heard from him, he was trying to track >> down an ITBLL issue that shows at scale. I've been trying to repro at >> smaller scale. I think that is what is going on (Mikhail fix me where I >> have it wrong). >> >> St.Ack >> >> On Wed, Aug 3, 2016 at 8:58 PM, Nick Dimiduk <[email protected]> wrote: >> >> > Hi folks, >> > >> > Just curious on the status of 1.3 release. We were close but momentum >> has >> > slowed? Is there someplace where this is tracked? Looks like the items >> > listed in HBASE-15341 are wrapped up. >> > >> > Thanks, >> > Nick >> > >> > > > > -- > Thanks, > Michael Antonov >
