Please wait. Fix for PHOENIX-2649 doesn't work as expected. So, I reopened it and attached a new fix. For PHOENIX-1973 there will be some performance improvement and I need to retest it carefully with applied PHOENIX-2649.
Thanks, Sergey On Wed, Feb 3, 2016 at 6:44 PM, James Taylor <[email protected]> wrote: > Ok, let's roll a new RC then. PHOENIX-2649 has a fix as do a few others. > Let's see if we can get a fix in for PHOENIX-2647 before cutting a new RC. > If PHOENIX-1973 is ready, that one is fine too IMHO. > > On Tue, Feb 2, 2016 at 11:56 PM, Gabriel Reid <[email protected]> > wrote: > >> PHOENIX-1973 doesn't seem like a blocker to me, although it would >> certainly be a good improvement. >> >> PHOENIX-2649 does look like a blocker however -- I think that the >> potential issues for an incorrect WritableComparator in an MR job >> could likely include doing the wrong thing if an OOME isn't triggered. >> >> - Gabriel >> >> On Wed, Feb 3, 2016 at 1:07 AM, James Taylor <[email protected]> >> wrote: >> > Are PHOENIX-1769, PHOENIX-1973, or PHOENIX-2649 blockers for 4.7.0 >> release? >>
