Zeebe 0.23.4

Release: 0.23.4

Bug Fixes

Broker

• Dispatcher lets you write batches that are larger than "maxFragmentLength" (#4851)
• Elasticsearch exporter bulk is not limited (#4668)
• No fail over when stream process fails (#4723)
• An error event is not caught by a boundary event on a multi-instance subprocess (#4601)
• Follower partitions are falsely marked as unhealthy (#4618)

Gateway

• Long Polling is blocked even though jobs are available on the broker (#4396)

Merged Pull Requests

Pull requests

• [BACKPORT] fix(dispatcher): don't allow claiming more than max size (#4863)
• [BACKPORT] fix(broker): use logstream name for health monitoring (#4843)
• [BACKPORT] Probe components health when partition is probed (#4796)
• [BACKPORT] fix(exporters/elasticsearch-exporter): avoid OOM exception when retrying (#4790)
• [BACKPORT] fix(engine): catch error event on multi-instance activity (#4752)
• [BACKPORT] chore(broker): Keep records of active long polling requests (#4681)