Skip to content
Open
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
3 changes: 1 addition & 2 deletions docs/streams/upgrade-guide.md
Original file line number Diff line number Diff line change
Expand Up @@ -300,7 +300,7 @@ We added a new Serde type `Boolean` in [KIP-907](https://cwiki.apache.org/conflu

[KIP-770](https://cwiki.apache.org/confluence/pages/viewpage.action?pageId=186878390) deprecates config `cache.max.bytes.buffering` in favor of the newly introduced config `statestore.cache.max.bytes`. To improve monitoring, two new metrics `input-buffer-bytes-total` and `cache-size-bytes-total` were added at the DEBUG level. Note, that the KIP is only partially implemented in the 3.4.0 release, and config `input.buffer.max.bytes` is not available yet.

[KIP-873](https://cwiki.apache.org/confluence/pages/viewpage.action?pageId=211883356) enables you to multicast result records to multiple partition of downstream sink topics and adds functionality for choosing to drop result records without sending. The `Integer StreamPartitioner.partition()` method is deprecated and replaced by the newly added `Optiona≶Set<Integer>>StreamPartitioner.partitions()` method, which enables returning a set of partitions to send the record to.
[KIP-873](https://cwiki.apache.org/confluence/pages/viewpage.action?pageId=211883356) enables you to multicast result records to multiple partition of downstream sink topics and adds functionality for choosing to drop result records without sending. The `Integer StreamPartitioner.partition()` method is deprecated and replaced by the newly added `Optional<Set<Integer>> StreamPartitioner.partitions()` method, which enables returning a set of partitions to send the record to.

[KIP-862](https://cwiki.apache.org/confluence/x/WSf1D) adds a DSL optimization for stream-stream self-joins. The optimization is enabled via a new option `single.store.self.join` which can be set via existing config `topology.optimization`. If enabled, the DSL will use a different join processor implementation that uses a single RocksDB store instead of two, to avoid unnecessary data duplication for the self-join case.

Expand Down Expand Up @@ -592,4 +592,3 @@ compatible
* [Documentation](/documentation)
* [Kafka Streams](/documentation/streams)


Loading