Author: Ismael Juma <ismael@juma.me.uk>
Reviewers: Sriharsha Chintalapani <schintalapani@hortonworks.com>, Ben Stopford <benstopford@gmail.com>, Jun Rao <junrao@gmail.com>
Closes#273 from ijuma/kafka-2517-ssl-zero-copy-regression
Author: Parth Brahmbhatt <brahmbhatt.parth@gmail.com>
Author: Ismael Juma <ismael@juma.me.uk>
Reviewers: Ismael Juma <ismael@juma.me.uk>, Jun Rao <junrao@gmail.com>
Closes#230 from Parth-Brahmbhatt/KAFKA-2212
It compares upper bound with itself.
Author: Edward Ribeiro <edward.ribeiro@gmail.com>
Reviewers: Aditya Auradkar, Ismael Juma, Guozhang Wang
Closes#182 from eribeiro/equals-bug
They now both use 2 spaces for indents, which is what `checkstyle.xml` was
already doing. `import.xml` had a mixture of tabs and 4 spaces previously.
Author: Ismael Juma <ismael@juma.me.uk>
Reviewers: Gwen Shapira
Closes#253 from ijuma/fix-xml-indents
GC is set to G1 collector.
Author: Manikumar reddy O <manikumar.reddy@gmail.com>
Reviewers: Ismael Juma <ismael@juma.me.uk>, Jun Rao <junrao@gmail.com>
Closes#218 from omkreddy/KAFKA-2514
This work has been contributed by Jesse Anderson, Randall Hauch, Yasuhiro Matsuda and Guozhang Wang. The detailed design can be found in https://cwiki.apache.org/confluence/display/KAFKA/KIP-28+-+Add+a+processor+client.
Author: Guozhang Wang <wangguoz@gmail.com>
Author: Yasuhiro Matsuda <yasuhiro.matsuda@gmail.com>
Author: Yasuhiro Matsuda <yasuhiro@confluent.io>
Author: ymatsuda <yasuhiro.matsuda@gmail.com>
Author: Randall Hauch <rhauch@gmail.com>
Author: Jesse Anderson <jesse@smokinghand.com>
Author: Ismael Juma <ismael@juma.me.uk>
Author: Jesse Anderson <eljefe6a@gmail.com>
Reviewers: Ismael Juma, Randall Hauch, Edward Ribeiro, Gwen Shapira, Jun Rao, Jay Kreps, Yasuhiro Matsuda, Guozhang Wang
Closes#130 from guozhangwang/streaming
hachikuji ewencp I found this problem when adding new consumer to mirror maker which commits offset in the rebalance callback. It is not clear to me why we are triggering rebalance for commitSync() and fetchCommittedOffset(). Can you help review to see if I miss something?
Regarding commitSync, After each poll() the partitions will be either assigned to a consumer or it will be already revoked. As long as user is using internal offset map, the offset map will always be valid. i.e. the offset map will always only contain the assigned partitions when commitSync is called. Hence there is no need to trigger a rebalance in commitSync().
The same guarantee also apply to fetchCommittedOffset(), isn't the only requirement is to ensure we know the coordinator?
Another related issue is that today the IllegalGenerationIdException is a bit confusing. When we receive an IllegalGenerationIdException from heartbeat, we need to use that same generation Id to commit offset and the coordinator will take it. So the generation ID was not really illegal. I will file a ticket for this issue.
Author: Jiangjie Qin <becket.qin@gmail.com>
Reviewers: Jason Gustafson, Guozhang Wang
Closes#221 from becketqin/KAFKA-2555
Updated the version from 0.8.3 to 0.9.0. in ApiVersion. Also updated in gradle.propeties.
Author: Manikumar reddy O <manikumar.reddy@gmail.com>
Reviewers: Ismael Juma, Gwen Shapira
Closes#237 from omkreddy/KAFKA-2554
Simplified the logic to choose the default fix version. We just hardcode
it for `trunk` and try to compute it based on the branch name for the
rest.
Removed logic that tries to handle forked release branches as it
seems to be specific to how the Spark project handles their JIRA.
Author: Ismael Juma <ismael@juma.me.uk>
Reviewers: Guozhang Wang
Closes#238 from ijuma/kafka-2548-merge-pr-tool-4-segment-fix-version
This is a patch to get around the problem discussed in the KAFKA-1387 jira. The tests are not passing in my box when I run them all, but they do pass when I run them individually, which indicates that there is something leaking from a test to the next. I still need to work this out and also work on further testing this. I wanted to open this PR now so that it can start getting reviewed.
Author: flavio junqueira <fpj@apache.org>
Author: fpj <fpj@apache.org>
Author: Flavio Junqueira <fpj@apache.org>
Reviewers: Ismael Juma <ismael@juma.me.uk>, Guozhang Wang <wangguoz@gmail.com>, Jun Rao <junrao@gmail.com>
Closes#178 from fpj/1387
We now write to the channel with an empty buffer when
there are pending bytes remaining and all data has been
sent.
Author: Ismael Juma <ismael@juma.me.uk>
Reviewers: Sriharsha Chintalapani <schintalapani@hortonworks.com>, Ben Stopford <benstopford@gmail.com>, Jun Rao <junrao@gmail.com>
Closes#236 from ijuma/kafka-2576-ssl-multi-partition-topic-hang
Author: Dong Lin <lindong28@gmail.com>
Author: Dong Lin <lindong@cis.upenn.edu>
Reviewers: Jason Gustafson, Ismael Juma, Guozhang Wang
Closes#118 from lindong28/KAFKA-2390
Author: Parth Brahmbhatt <brahmbhatt.parth@gmail.com>
Reviewers: Ismael Juma <ismael@juma.me.uk>, Jun Rao <junrao@gmail.com>
Closes#195 from Parth-Brahmbhatt/KAFKA-2211
I have reopened this issue because the controller isn't cleaning up the state upon an exception and the test case was legitimately failing for me every now and then. I'm proposing a change to fix this.
Author: fpj <fpj@apache.org>
Author: flavio junqueira <fpj@apache.org>
Reviewers: Ismael Juma, Gwen Shapira, Guozhang Wang
Closes#212 from fpj/2300
Author: Ismael Juma <ismael@juma.me.uk>
Reviewers: Aditya Auradkar <aauradkar@linkedin.com>, Jun Rao <junrao@gmail.com>
Closes#194 from ijuma/kafka-2440-use-network-client-in-fetcher
A shot to remove commit type from new consumer. The coordinator constructor takes a default offset commit callback mainly for testing purpose.
Author: Jiangjie Qin <becket.qin@gmail.com>
Reviewers: Ewen Cheslack-Postava, Jason Gustafson, Guohang Wang
Closes#134 from becketqin/KAFKA-2389
Author: Ashish Singh <asingh@cloudera.com>
Reviewers: Jason Gustafson, Guozhang Wang, Edward Ribeiro, Ismael Juma
Closes#128 from SinghAsDev/KAFKA-1893
ewencp
The changes here are smaller than they look - mostly refactoring/cleanup.
- ConsumerPerformanceService: added new_consumer flag, and exposed more command-line settings
- benchmark.py: refactored to use `parametrize` and `matrix` - this reduced some amount of repeated code
- benchmark.py: added consumer performance tests with new consumer (using `parametrize`)
- benchmark.py: added more detailed test descriptions
- performance.py: broke into separate files
Author: Geoff Anderson <geoff@confluent.io>
Reviewers: Ewen Cheslack-Postava, Jason Gustafson, Gwen Shapira
Closes#179 from granders/KAFKA-2489-benchmark-new-consumer
Small clarification to docs. Current behaviour could confuse when doing something like:
consumer.seekToEnd()
consumer.send(msg)
consumer.poll() //would return msg as seek evaluates lazily
Author: Ben Stopford <benstopford@gmail.com>
Reviewers: Gwen Shapira
Closes#199 from benstopford/minor-stuff
Ismael Juma
Parth Brahmbhatt
Jason Gustafson
Edward Ribeiro
Mayuresh Gharat
Dong Lin
Manikumar reddy O
Ewen Cheslack-Postava
Ashish Singh
Guozhang Wang
Jiangjie Qin
flavio junqueira
Onur Karaman
Joel Koshy
Geoff Anderson
jholoman
Ben Stopford