Skip to content

Latest commit

 

History

History
110 lines (80 loc) · 4.11 KB

prev_releases.adoc

File metadata and controls

110 lines (80 loc) · 4.11 KB

Past Releases

  • 0.33

    • Added support for embedding Forklift Server and runtime deployments

  • 0.32

    • Improve performance of replay indexing.

  • 0.31

  • 0.30

    • Add ordering of messages. Ensures exclusive consumers process messages based on an abitrary order specified by the consumer.

  • 0.29

    • Improve ES indexing performance by using the transport client rather than REST.

  • 0.28

    • Performance fix for Replay ES. Shouldn’t block all threads based on the indexing resp time of elastic search.

    • Correct shutdown issue that caused a queue to be completely consumed before allowing forklift to shutdown.

  • 0.27

    • Log errors from messages that will be retried as warnings.

  • 0.26

    • Startup deployment scanning order changed to avoid a race condition where a property value may not have been loaded before the start of a consumer.

    • Fix validation issue that caused lists to not be validated correctly.

  • 0.25

    • Fix multithreading that didn’t scale past two threads.

    • Include bash in the alpine based docker container so work arounds aren’t necessary to start the system.

  • 0.24

    • Synchronous producers

  • 0.23

  • 0.22

    • Add retries to es indexing code for the retry and replay plugins.

  • 0.21

    • Ensure consumers connections are cleaned up correctly.

  • 0.20

    • Fix thread pool restart after a consumer receives a badly formed message.

  • 0.18

    • Acknowledge messages once forklift has obtained the message and begins processing it. This corrects a bug that could cause a message to be processed twice in the event of a broker node failure.

  • 0.17

    • Add a separate thread to push replay logs to elastic search. This will allow msgs to be unhindered by the REST calls to ES.

    • Cleanup old errors from previous indexes before indexing the new doc. This can be inefficient, but guarantees the msg id is only in the system once.

  • 0.16

    • Fix bug that causes the max retries exceeded prop to not show in replay logs.

    • Fix bug that caused the retry startup search to fail due to having too many results.

  • 0.15

    • Elasticsearch implementation of retries and replay logs.

  • 0.13

    • Fix MultiThreaded consumer auto closing of resources.

  • 0.12

    • Auto closing of producers that are injected.

  • 0.11

    • Close jms sessions correctly to avoid open session buildup on the amq nodes.

  • 0.10

    • Added a new @Property annotation that injects message properties into the consumer

    • Extended @Config and @Headers to allow you to inject specific headers and config fields

    • The @On decorator now is repeatable; one method can now intercept multiple lifecycle events

    • Override classloader to delegate to the JVM libraries in cases where libs that start with these packages ("java", "com.sun", "sun", "org.xml", "org.w3c").

  • 0.9

    • Removed Scala build dependencies.

    • Added the ability to close the forklift producer to avoid memory leaks in client code.

    • Added key/value pair send viathe forklift producer.

    • Added a new On decorator that allows the lifecycle to be intercepted at any point by a method intercept.

  • 0.8

    • Remove spring completely from the core and replace with a service mechanism.

  • 0.7

    • Fix issue that would cause a consumer to stop processing due to a redelivery loop. This was caused any time a bad json parse occurred.

  • 0.6

    • Replay logger flush periodically if no messages are received. Helps guarantee nothing gets lost in the buffer.

    • Log file configuration. Setup the container to drop a log file in /logs.

    • Fix lost msgs that end up in the dead letter queue by using the core thread to process messages during a thread pool overload.

    • Allow multiple Queue/Topic annotations per consumer. This helps when needing to monitor two different queues for fast vs slow processing.

  • 0.5

    • Spring setup fixes

    • Main thread death fix when a consumer’s spring context startup fails

  • 0.4

    • Allow the consul host to be specified

    • Fix failover url construction when using service discovery with only one broker.

  • 0.3

    • Updates to docker container builds

    • New arguments to allow retry/replay dirs to be set

    • Allow two deployment directories to be scanned

  • 0.2

    • Fix for property file injection into consumer.

  • 0.1

    • Initial beta release