The 1.1.9 release is the ninth bugfix release for the 1.1 series. This release is a smaller bugfix release, which has two important changes. The first one is that as of now, passing null for both bucket and bucket password will immediately throw a proper exception.
The other bugfix makes sure that during ReplicateTo and PersistTo phases, always the latest cluster config is used. This prevents an issue where the topology changes during the "observe" loop and exceptions are thrown because of data mismatch.
When using the TapClient to stream data through the tapBackfill operation, the code now silently ignores data that arrives from replica nodes. This is a very basic deduplication mechanism and will be enhanced in the future to also reduce the amount of network traffic.
As of 1.1.9, when null is passed in for either the bucket name or the bucket password, a proper exception is thrown immediately. This "fail fast" mechanism will make it more clear what is going on during a failed startup procedure.
When using the PersistTo and/or ReplicateTo settings, the code now - during the "observe loop" - always fetches an up-to-date configuration from the provider. This helps to mitigate issues where the cluster configuration changes during such an interval which potentially leads to exceptions.