Re: [VOTE] First release candidate for HBase 1.1.3 (RC0) is available

For anyone who's not following along on HBASE-14944, I added a comment based on the fine explanation provided by Andrew and the wording in our book. Unless there are objections, I will cary on with the release, including these changes. If you do have an objection, please speak up. Thanks, Nick

Re: [VOTE] First release candidate for HBase 1.1.3 (RC0) is available

So it's either EOL 1.0 (already intended) and 1.1 due to these security bugs, or ship a source-incompatible patch release? The classes are all marked LimitedPrivate("coprocessors"), so the scope it's as wide as it could be. Is it not possible to provide default implementations of the new methods?

Re: [VOTE] First release candidate for HBase 1.1.3 (RC0) is available

> ​ So it's either EOL 1.0 (already intended) and 1.1 due to these security bugs, or ship a source-incompatible patch release? I opened HBASE-14944 for this > Is it not possible to provide default implementations of the new methods? We did. The addition of the new methods to the interfaces is

Re: [VOTE] First release candidate for HBase 1.1.3 (RC0) is available

> Reopened HBASE-14605, HBASE-14655, and HBASE-14631 as blockers for 1.1.x and 1.0.x. I don't think we could leave the underlying problems with incorrect user context for security alone on branches 1.0 and 1.1 unfixed. We can revert this stuff - these changes are at the end of a cascade of

Re: [VOTE] First release candidate for HBase 1.1.3 (RC0) is available

I found some source-incompatible changes in the compatibility report [0]. Reopened HBASE-14605, HBASE-14655, and HBASE-14631 as blockers for 1.1.x and 1.0.x. Since I'm asking, would be good to see the last mile on HBASE-14822 and HBASE-14844. [0]:

Re: [VOTE] First release candidate for HBase 1.1.3 (RC0) is available

With HBASE-14689 resolved, I plan to spin the next RC this weekend. Let me know if you have any concerns. -n On Tue, Nov 17, 2015 at 8:33 PM, Andrew Purtell wrote: > Let's revert this from 0.98 even if the problem isn't manifesting there > (or perhaps just not in the

Re: [VOTE] First release candidate for HBase 1.1.3 (RC0) is available

-1. I think this RC is also affected by HBASE-14689: A run of bin/hbase pe --latency --nomapred --presplit=10 --valueSize=10 randomWrite 10 results in client hang and a lot of following exceptions. Interesting that, I cannot reproduce this with LTT. java.io.IOException: Timed out waiting

Re: [VOTE] First release candidate for HBase 1.1.3 (RC0) is available

Okay, good find Enis. On Tuesday, November 17, 2015, Enis Söztutar wrote: > Sorry for the trouble, let me revert the commit. I'll look at the 0.98 RC > as well. > > > Enis > > On Tue, Nov 17, 2015 at 7:32 PM, Enis Söztutar > wrote: > > >

Re: [VOTE] First release candidate for HBase 1.1.3 (RC0) is available

Let's revert this from 0.98 even if the problem isn't manifesting there (or perhaps just not in the same way) > On Nov 17, 2015, at 7:33 PM, Enis Söztutar wrote: > > Sorry for the trouble, let me revert the commit. I'll look at the 0.98 RC > as well. > > > Enis > >> On

Re: [VOTE] First release candidate for HBase 1.1.3 (RC0) is available

Indeed interesting that LTT doesn't catch it. > On Nov 17, 2015, at 7:32 PM, Enis Söztutar wrote: > > -1. > > I think this RC is also affected by HBASE-14689: > > > A run of > > bin/hbase pe --latency --nomapred --presplit=10 --valueSize=10 randomWrite > 10 > >

Re: [VOTE] First release candidate for HBase 1.1.3 (RC0) is available

Sorry for the trouble, let me revert the commit. I'll look at the 0.98 RC as well. Enis On Tue, Nov 17, 2015 at 7:32 PM, Enis Söztutar wrote: > -1. > > I think this RC is also affected by HBASE-14689: > > > A run of > > bin/hbase pe --latency --nomapred --presplit=10

Re: [VOTE] First release candidate for HBase 1.1.3 (RC0) is available

+1 Checked signature and sums Unpacked binary and source tarballs, contents look good Built from source RAT check passes Loaded 1M rows with LoadTestTool, no unexpected log messages, reported latencies in line with expectations Ran IntegrationTestBigLinkedList, no errors On Sun, Nov 8, 2015 at

Re: [VOTE] First release candidate for HBase 1.1.3 (RC0) is available

+1 Everything looks to be in the correct location. Spun up a regionserver inserted some things. On Sun, Nov 15, 2015 at 7:38 PM, Nick Dimiduk wrote: > The issue is not in the product but in the report generation. It seems the > reporting tool does not work correctly on my

Re: [VOTE] First release candidate for HBase 1.1.3 (RC0) is available

What is the specific compatibility issue? By default the tool will only look at classes we annotate as Public or LimitedPrivate. Maybe that's why some changes you are expecting to be called out are not present in the report? > On Nov 15, 2015, at 3:07 PM, Nick Dimiduk

Re: [VOTE] First release candidate for HBase 1.1.3 (RC0) is available

For my part, my vote is +0, pending review of compatibility report. My vote is +1 assuming we can get a clean report. - download tarballs, verified signatures vs below ndimiduk.asc and KEYS file in dist repo - verified none of rej, orig, java, or scala files in the bin tgz - verified none of rej,

Re: [VOTE] First release candidate for HBase 1.1.3 (RC0) is available

The issue is not in the product but in the report generation. It seems the reporting tool does not work correctly on my new hardware. I ran it on another machine and this one looks correct. I've updated the file, at the same address [0]. This report shows a number of added methods, but still

Re: [VOTE] First release candidate for HBase 1.1.3 (RC0) is available

Thank you Josh for taking time to evaluate this release candidate. A reminder to others that the voting period is scheduled to expire in roughly 24 hours. I myself have not had time this week to evaluate this candidate properly, so I would like to extend the voting period through the weekend. If

Re: [VOTE] First release candidate for HBase 1.1.3 (RC0) is available

+1 (non-binding) * Built from source * Ran tests (-PrunDevTests). o.a.h.h.r.TestRegionServerHostname was problematic, might have just been me. * Checked sigs/xsums * Checked the compat report (thanks for posting it, Nick) * Skimmed release notes looking for anything that might introduce new

[VOTE] First release candidate for HBase 1.1.3 (RC0) is available

I'm happy to announce the first release candidate of HBase 1.1.3 (HBase-1.1. 3RC0) is available for download at https://dist.apache.org/repos/dist/dev/hbase/hbase-1.1.3RC0/ Maven artifacts are also available in the staging repository