This page describes what works and what does not with the current state of the repository.
Compatibility is aiming Apache Kafka 0.10.2.1
- The API parsing front-end. Mostly done. Support for new calls and their versions can be added with a few lines of code
- Producing to topics
- Consuming from topics (HEAD, beginning, etc)
- Compacted topics
- Data cleanup thread
- Compression
- Consumer groups. Currently any consumer connecting is being told it is the only consumer in any group.
- Default Kafka Java client
- librdkafka (current master, tested on 7a2b80d5daeafca99c852b0406d64edc7267aefb)
- various older versions
- Custom backend with high throughput write support.
- Server stabiliy code (don't panic so much on errors)
- Protocol extensions to make performance better (e.g. empty response being empty)
- Admin UI
- track producers
- track consumers
# Default java client tested like this:
./bin/kafka-console-producer.sh --broker-list 127.0.0.1:9092 --topic test --property parse.key=true --property key.separator=:
./bin/kafka-console-consumer.sh --bootstrap-server 127.0.0.1:9092 --topic test --from-beginning --property print.key=true
./bin/kafka-console-consumer.sh --bootstrap-server 127.0.0.1:9092 --topic test --property print.key=true
# librdkafka tested using the latest build of kafkacat on top of latest build of the library
./kafkacat -P -b 127.0.0.1 -t test -c 2
./kafkacat -C -b 127.0.0.1 -t test
./kafkacat -L -b 127.0.0.1
./kafkacat -L -b 127.0.0.1 -t test
./kafkacat -Q -b 127.0.0.1 -t test:0:-1 # -1 is the latest offset
./kafkacat -Q -b 127.0.0.1 -t test:0:-2 # -2 is the earliest available offset
./kafkacat -b 127.0.0.1 -G gr1 test