Meh... Looked through the test history, you're right. I'll back it out. This is the 2nd time we're backing this out. I will remove this from 0.94 altogether.
-- Lars ________________________________ From: lars hofhansl <[email protected]> To: Ted Yu <[email protected]>; "[email protected]" <[email protected]> Sent: Thursday, March 14, 2013 2:32 PM Subject: Re: no 0.94 commits please, for 0.94.6RC2 I ran the tests locally multiple time and it was fine. I do not think we need to back this out... Again. ________________________________ From: Ted Yu <[email protected]> To: [email protected]; lars hofhansl <[email protected]> Sent: Thursday, March 14, 2013 2:27 PM Subject: Re: no 0.94 commits please, for 0.94.6RC2 Lars: Jeffrey, Devaraj and I worked together on the test failure of TestMasterFailover#testMasterFailoverWithMockedRITOnDeadRS in 0.94 This test passed pretty consistently before HBASE-7824 went in. However, it fails quite frequently in current 0.94 branch. Jeffrey had some initial finding which unfortunately didn't match the test output we looked at. We will continue to find out the relationship between HBASE-7824 and TestMasterFailover#testMasterFailoverWithMockedRITOnDeadRS failure. Meanwhile, to unblock 0.94.6 RC2, I plan to back out HBASE-7824. Cheers On Thu, Mar 14, 2013 at 2:15 PM, lars hofhansl <[email protected]> wrote: I am still waiting for any jenkins run. Decided to build the release myself (running through all tests, just like jenkins would). > > > >________________________________ > From: lars hofhansl <[email protected]> >To: hbase-dev <[email protected]> >Sent: Thursday, March 14, 2013 11:16 AM >Subject: no 0.94 commits please, for 0.94.6RC2 > > >Waiting for a successful jenkins run.
