0.7.0
Enhancements
- Log file is now world readable and credentials are redacted
- New dedicated Documentation Website
- Support Choria Organization Issuer based Choria Brokers
- Move some noisy log lines to debug level
- Enable profiling on the monitor port
- Support being deployed in High Available mode with active-active multi node reliable clusters
- Add numerous CLI tools for inspecting cluster traffic, state files and search advisories
- Support more template fields in advisory subjects
- Adds a new replication mode optimised for high-throughput under high latency conditions
- Synchronise ID track state using a gossip protocol over the NATS cluster
- Increase backoff window during long outages
Removals
- Remove the stream lag metric that was confusing and difficult to use in alerts and graphs
Bug Fixes
- Authentication errors from the brokers will not cause the reconnection cycle to terminate
- Fix health management routines that would consume 100% cpu on one core in some cases
- Handle situations where the target or source stream was purged
Full Changelog: v0.6.1...v0.7.0