The Elasticsearch Logstash team is pleased to announce the release of version 1.4.1. This is a minor version containing mostly bugfixes and minor feature enhancements. This version of Logstash now contains an embedded version of Elasticsearch 1.1.1, as well as Kibana 3.0.1.
We strive to constantly improve all aspects of our software, both glorious and mundane. To that end, we have updated our documentation, which, as always, is driven by our community of users - both via feedback and actual user documentation submissions. Also, the project's testing infrastructure received some love in this cycle, as our team has been busy improving the specs and tests as well as our continuous integration system. Lastly, the packaging has been greatly improved, ensuring users on all of our supported platforms have great experiences when deploying Logstash!
Inputs, Outputs and Filters
- The multiline filter received a makeover, fixing some bugs and making it overall more awesome. In particular, the infamous 'tv_sec' bug was fixed.
- The Twitter input now allows access to full Twitter tweet data, by setting (not surprisingly) the `full_tweet` option.
- Updated RabbitMQ input/output library which fixes a threading leak, and decreases latency.
- Grokking your data just got more powerful, with the addition and/or modification of built-in patterns!
- The embedded elasticsearch output debuted a couple fixes which should eliminate errors which many users have experienced ("waited for 30s..."). Wait no more!
- The collectd codec is new and supersedes the older collectd input plugin. When used in conjunction with our updated UDP input, the performance has increased by over 15x.
- Created a more generic init script created by PleaseRun.
- Fixed core and contrib dependencies so contrib can be installed.
- For future releases we've implemented nightly package install tests.