Jay Kreps
|
e7edb5e1e9
|
KAFKA-631 Implement a log cleaner for Kafka. Reviewed by Neha.
|
12 years ago |
Sriram Subramanian
|
40a80fa7b7
|
KAFKA-700 log client ip when we log each request on the broker; reviewed by Neha Narkhede
|
12 years ago |
Sriram Subramanian
|
8d41620a42
|
KAFKA-696 Fix toString() API for all requests to make logging easier to read; reviewed by Neha Narkhede, Jun Rao
|
12 years ago |
Neha Narkhede
|
a15f1f2d84
|
KAFKA-720 Migration tool halts silently; reviewed by Neha Narkhede
|
12 years ago |
Neha Narkhede
|
8e7effc055
|
Minor log4j fix in the producer
|
12 years ago |
Jun Rao
|
7833c894a8
|
KAFKA-714: ConsoleConsumer throws SocketTimeoutException when fetching topic metadata; reviewed by Neha Narkhede
|
12 years ago |
Edward Jay Kreps
|
311a5d81d8
|
KAFKA-727 broker can still expose uncommitted data to a consumer; reviewed by Neha Narkhede
|
12 years ago |
Neha Narkhede
|
eb0d5a7f81
|
KAFKA-708 ISR becomes empty while marking a partition offline; reviewed by Jun Rao
|
12 years ago |
Jun Rao
|
307b1f338a
|
Add ReplicaFetcherThread name to mbean names; kafka-726; patched by Swapnil Ghike; reviewed by Jun Rao
|
12 years ago |
Joel Koshy
|
0d197f9b97
|
KAFKA-705 Shutting down brokers should not receive stop replica requests if controlled shutdown status is incomplete; reviewed by Neha Narkhede
|
12 years ago |
Joe Stein
|
5e3e181ace
|
KAFKA-139 cross-compile multiple Scala versions the dependency jars moved and the bin scripts needed to point to the new locations
|
12 years ago |
Joel Koshy
|
f112ef2b57
|
KAFKA-705 Allow restarted replicas to follow a leader that is on a shutting down broker; reviewed by Neha Narkhede.
|
12 years ago |
Neha Narkhede
|
5490884e9c
|
Revert "draft patch"
This reverts commit aa1546b090 .
|
12 years ago |
Neha Narkhede
|
426ef16625
|
KAFKA-707 Improve error message in the producer when sending data to a partition without an active leader; reviewed by Jun Rao
|
12 years ago |
Joe Stein
|
1e011bf42e
|
KAFKA-709 change in default blocking tripped test checking queue is full so updated test in case someone changes the new default to something else
|
12 years ago |
Neha Narkhede
|
aa1546b090
|
draft patch
|
12 years ago |
Neha Narkhede
|
ecd1f19f5a
|
KAFKA-709 Default queue.enqueue.timeout.ms to -1; reviewed by Jay Kreps
|
12 years ago |
Jun Rao
|
214a0af46b
|
Consumer rebalance fails if no leader available for a partition and stops all fetchers; patched by Maxime Brugidou; reviewed by Jun Rao; kafka-693
|
12 years ago |
Edward Jay Kreps
|
da7f14676e
|
KAFKA-702 Deadlock between request handler/processor threads; reviewed by Neha Narkhede, Jun Rao
|
12 years ago |
Jun Rao
|
777f662201
|
Disallow clients to set replicaId in FetchRequest; kafka-699; patched by Jun Rao; reviewed by Neha Narkhede
|
12 years ago |
Jay Kreps
|
de1a4d7276
|
KAFKA-698 Avoid advancing the log end offset until the append has actually happened since reads may be happening in the meantime.
|
12 years ago |
Swapnil Ghike
|
eeb817ac23
|
KAFKA-697 ConsoleConsumer throws InvalidConfigException for . in client id; reviewed by Jun Rao, Neha Narkhede, John Fung
|
12 years ago |
Jun Rao
|
b2eec991d3
|
ConsoleConsumer throws InvalidConfigException; kafka-697; patched by Swapnil Ghike; reviewed by Jun Rao
|
12 years ago |
Jun Rao
|
53818bb7ee
|
ConsoleProducer does not exit correctly; kafka-701; patched by Maxime Brugidou; reviewed by Jun Rao
|
12 years ago |
Jun Rao
|
a40953196e
|
Use uniform convention for naming properties keys; kafka-648; patched by Sriram Subramanian; reviewed by Jun Rao
|
12 years ago |
Neha Narkhede
|
03eb903ce2
|
KAFKA-683 Fix correlation id in all requests sent to kafka; reviewed by Jun Rao
|
12 years ago |
Ben Fleis
|
c12608c5fe
|
KAFKA-692 ConsoleConsumer outputs diagnostic message to stdout instead of stderr; reviewed by Neha Narkhede
|
12 years ago |
Jun Rao
|
b71e6dc352
|
Fault tolerance broken with replication factor 1; kafka-691; patched by Maxime Brugidou; reviewed by Jun Rao
|
12 years ago |
Neha Narkhede
|
e0b3b63163
|
KAFKA-690 TopicMetadataRequest throws exception when no topics are specified, reviewed by Jay Kreps and Neha Narkhede
|
12 years ago |
Jun Rao
|
85c9e91c80
|
ConsoleProducer does not have the queue-size option; kafka-684; patched by Maxime Brugidou; reviewed by Jun Rao
|
12 years ago |
Jun Rao
|
fd94251d89
|
ConsumerOffsetChecker does not work with 0.8; kafka-685; patched by Maxime Brugidou; reviewed by Jun Rao
|
12 years ago |
Jay Kreps
|
274cd02f1a
|
KAFKA-657 Change property name to offset.metadata.max.bytes for consistency.
|
12 years ago |
David Arthur
|
222c0e46ab
|
KAFKA-657 Add APIs for the consumer to commit and fetch offsets on the broker.
|
12 years ago |
Jun Rao
|
2f4bfc645a
|
ApiUtils#writeShortString uses String length instead of byte length; kafka-680; patched by David Arthur; reviewed by Jun Rao
|
12 years ago |
Jun Rao
|
6cf46e7cab
|
High watermark recovered incorrrectly from file; kafka-681; patched by Jun Rao; reviewed by Joel Koshy, Neha Narkhede and John Fung
|
12 years ago |
Joel Koshy
|
25d77cc69c
|
KAFKA-668 Store jmx port in broker zk string - controlled shutdown admin tool should not require controller JMX url/port to be supplied. Reviewed by Jun Rao.
|
12 years ago |
Jay Kreps
|
c6d41102d8
|
KAFKA-675 Allow the user to override the host that we bind to. Patch from Matan Amir<matan.amir@voxer.com> with slight changes
to improve error messages for a bad host or port.
|
12 years ago |
Joel Koshy
|
04b52743f8
|
KAFKA-664 RequestPurgatory should clean up satisfied requests from watchers map. Also, simplify the purge logic - purge based on an incoming request interval.
|
12 years ago |
Swapnil Ghike
|
994fe081b6
|
KAFKA-646 Missing files from previous commit
|
12 years ago |
Jun Rao
|
af83dfbe0c
|
trivial change to increase default offset commit interval to reduce ZK load
|
12 years ago |
Swapnil Ghike
|
8ad9018ea1
|
KAFKA-646 Provide aggregate stats at the high level Producer and ZookeeperConsumerConnector level; reviewed by Neha Narkhede
|
12 years ago |
Jay Kreps
|
739b2df680
|
KAFKA-374 Move off the default java crc implementation to the crc code borrowed from Hadoop. Patch reviewed by Joe.
|
12 years ago |
Jay Kreps
|
f0d26335f2
|
KAFKA-673 The logic that controls whether to run recovery on the log is reversed. Reviewed by Neha.
|
12 years ago |
Derek Chen-Becker
|
45cb03d456
|
Re-applied old patch from KAFKA-139
Had to make some additional changes based on further mainline work. Leaving
KafkaProject.scala in place as a reference for now.
|
12 years ago |
Edward Jay Kreps
|
3244bcafe4
|
KAFKA-670 Clean spurious .index files; reviewed by Neha Narkhede
|
12 years ago |
Edward Jay Kreps
|
ac35ae1fef
|
KAFKA-636 Make log segment delete an asynchronous background action done by the scheduler. Patch reviewed by Jun and Neha.
git-svn-id: https://svn.apache.org/repos/asf/kafka/trunk@1420361 13f79535-47bb-0310-9956-ffa450edef68
|
12 years ago |
Edward Jay Kreps
|
f9702c66b7
|
KAFKA-667 Change the .highwatermark file name.
git-svn-id: https://svn.apache.org/repos/asf/kafka/branches/0.8@1420356 13f79535-47bb-0310-9956-ffa450edef68
|
12 years ago |
Edward Jay Kreps
|
a10be4b69c
|
KAFKA-597 Refactor scheduler. Fixes a couple of bugs, and adds the ability to mock scheduled tasks.
git-svn-id: https://svn.apache.org/repos/asf/kafka/trunk@1419692 13f79535-47bb-0310-9956-ffa450edef68
|
12 years ago |
Neha Narkhede
|
133c88278d
|
KAFKA-654 Irrecoverable error while trying to roll a segment that already exists; patched by Neha Narkhede; reviewed by Jay Kreps
git-svn-id: https://svn.apache.org/repos/asf/kafka/branches/0.8@1419627 13f79535-47bb-0310-9956-ffa450edef68
|
12 years ago |
Edward Jay Kreps
|
ab2494470c
|
KAFKA-642 Missed message magic version on last checkin.
git-svn-id: https://svn.apache.org/repos/asf/kafka/branches/0.8@1418423 13f79535-47bb-0310-9956-ffa450edef68
|
12 years ago |