I would love to see FLUME-2994 getting into the release (low-hanging fruit).
On Tue, Oct 4, 2016 at 6:01 PM, Balazs Donat Bessenyei <bes...@cloudera.com> wrote: > As there have been no objections, I am going to proceed with the plan > I have outlined in my original mail. > > I will try to work a little more on some flaky tests to help with the > release process. (Such as FLUME-3002: Some tests in TestBucketWriter > are flaky.) > > It would be awesome if someone could fix FLUME-2971 (Document Kerberos > set-up for Kafka Sink and Kafka Source). > > Also, I have found a few more tickets that could use a little care: > FLUME-2689 (reloading conf file leads syslogTcpSource not receives any > event) - probably only needs some more reviews, > FLUME-2716 (File Channel cannot handle capacity Integer.MAX_VALUE) - > missing a test, > FLUME-2461 (memoryChannel bytesRemaining counting error) - missing test > Patch-less (Needs some investigation and a fix. I am not sure about > their complexity): > FLUME-2912 (thrift Sources/Sinks can only authenticate with kerberos > principal in format with hostname) > FLUME-2871 (avro sink reset-connection-interval cause > EventDeliveryException) > FLUME-2811 (Taildir source doesn't call stop() on graceful shutdown) > > However, none of these seem to be a release blocker, so if they are > not done by Oct 7-10, it would still be fine. > > > Please, let me know your thoughts > > Donat > > On Mon, Oct 3, 2016 at 3:23 PM, Mike Percy <mpe...@apache.org> wrote: > > On Fri, Sep 30, 2016 at 7:01 PM, Saikat Kanjilal <sxk1...@hotmail.com> > > wrote: > > > >> I'd be willing to help, I've been heads down on other stuff and have had > >> to postpone the graph sink implementation (https://issues.apache.org/ > >> jira/browse/FLUME-2035) but am looking to resume work and targeting an > >> initial implementation by mid November sometime. In the meantime let me > >> know how I can get more deeply involved in the next release. > >> > > > > I'd recommend taking a look at things like documentation and filling in > > gaps so that we don't release with undocumented stuff. Other types of > > polish seem like a helpful thing to do right before a release as well. > > > > Mike >
