Thank you for the reviews. @Becket @Ewen, Agree that making all blocking calls have a timeout will be trickier and hence the scope of this KIP is limited to close().
@Jay Yes, this should definitely go into release notes, will make sure it is added. I will add some integration tests with broker failures for testing the timeout, but they cannot completely eliminate the risk of a hang. Over time, hopefully system tests will help catch most issues. On Sat, Dec 17, 2016 at 1:15 AM, Jay Kreps <j...@confluent.io> wrote: > I think this is great. Sounds like one implication is that existing code > that called close() and hit the timeout would now hang indefinitely. We saw > this kind of thing a lot in automated testing scenarios where people don't > correctly sequence their shutdown of client and server. I think this is > okay, but might be good to include in the release notes. > > -jay > > On Thu, Dec 15, 2016 at 5:32 AM, Rajini Sivaram <rsiva...@pivotal.io> > wrote: > > Hi all, > > > > > > I have just created KIP-102 to add a new close method for consumers with a > > > timeout parameter, making Consumer consistent with Producer: > > > > > > https://cwiki.apache.org/confluence/display/KAFKA/KIP- > 102+-+Add+close+with+timeout+for+consumers > > > > > > Comments and suggestions are welcome. > > > > > > Thank you... > > > > > > Regards, > > > > > > Rajini > -- Regards, Rajini
