Skip to content

Releases: harlow/kinesis-consumer

v0.3.6

18 Sep 06:39
4390050
Compare
Choose a tag to compare

Most notable changes are:
switch to AWS SDK2
adding support for split shards
bugfixes

v0.3.5

07 Sep 20:51
8257129
Compare
Choose a tag to compare

Final cut before moving to AWS SDK v2

Fast forward to current main line

25 Jun 23:41
400ef07
Compare
Choose a tag to compare

Haven't cut a release in a LONG time and there are several features (such as ScanInterval) and bugs fixed since v0.3.3

Fix dependency errors

31 Aug 20:21
Compare
Choose a tag to compare

Final stable release before adding ConsumerGroups

29 May 02:47
Compare
Choose a tag to compare

This will be the last Tagged release before adding consumer groups: #42

While the intention will be to keep master stable at all times, there are going to be some breaking changes around naming Checkpoint -> Storage and new Group interface.

Last release with ScanStatus

18 Feb 18:34
Compare
Choose a tag to compare

Going forward the concept of ScanStatus will be removed from the codebase. This will act as the final release before the breaking change is introduced into main branch.

There are a few bug fixes in this release too:

  • Moving cancellation of consumers into the error block (previously if a shard was closed all consumers would be signaled to shutdown)

Remove concept of client from consumer library

28 Dec 18:52
Compare
Choose a tag to compare

Having an additional Client has added some confusion (#45) on how to provide a
custom kinesis client. Allowing WithClient to accept a Kinesis client
it cleans up the interface.

Major changes:

  • Remove the Client wrapper; prefer using kinesis client directly
  • Change ScanError to ScanStatus as the return value isn't necessarily an error

Minor changes:

  • Additional test coverage

Last stable version with consumer/client split

29 Jul 05:17
Compare
Choose a tag to compare
Add unit tests for the GetRecords function (#64)

A closed shard returns the last sequence number and no error. The current implementation leads to an infinite loop if the shard is closed. NextShardIterator checking is enough. That's why I remove the getShardIterator call