4.3 Release Notes and Changelog (#2577)

Also remove stale attribution from NOTICE

NOTICE

@@ -1,5 +1,2 @@
-The class com.hazelcast.jet.impl.TopologicalSorter contains code
-derived from the Apache Tez project.
-
 The class com.hazelcast.jet.kafka.impl.ResumeTransactionUtil contains
-code derived from the Apache Flink project.
+code derived from the Apache Flink project.

hazelcast-jet-distribution/src/root/changelog.txt

@@ -1,3 +1,21 @@
+** Hazelcast Jet 4.3 **
+
+* @MohamedMandouh implemented distributed sorting: `BatchStage.sort()`.
+* @guenter-hesse added a high-throughput source with precisely controlled
+  timing, to be used to benchmark Jet's throughput and latency.
+* We added the ability for Jet to suspend an unbounded, fault-tolerant job
+  when it fails due to an exception in the pipeline code. This allows you
+  to diagnose the problem, remove its cause, and resume the pipeline without
+  data loss.
+* We have improved the consistency of reconnect behaviour across CDC sources and
+  introduced a uniform API to configure the reconnect strategy.
+* We improved the behavior of streaming pipelines based on ingestion time: a
+  low event rate in any partition of the source used to stall the progress,
+  now the watermark advances even without events.
+* We moved the `Pipeline` to `DAG` transformation from the client to the server,
+  in preparation for the upcoming feature of preventing event reordering in the
+  pipeline.
+
 ** Hazelcast Jet 4.2 **
 
 * We added Change Data Capture (CDC) support for PostgreSQL and MySql. New
@@ -39,7 +57,6 @@
   directly instead of requiring custom code in the `mapToOutputFn`. Credit
   for this one goes to @TomaszGaweda.
 
-
 ** Hazelcast Jet 4.0 **
 
 * Updated Hazelcast IMDG version to 4.0, which brings many new features

hazelcast-jet-distribution/src/root/release_notes.txt

@@ -23,6 +23,70 @@ To expand an issue number into a full GitHub link, use this template:
 
 https://github.com/hazelcast/hazelcast-jet/issues/<issue number>
 
+
+** Hazelcast Jet 4.3 **
+
+Members of the open source community that appear in these release notes:
+
+@caioguedes
+@guenter-hesse
+@MohamedMandouh
+
+Thank you for your valuable contributions!
+
+1. New Features
+
+[pipeline-api] [014] @MohamedMandouh implemented distributed sorting: `BatchStage.sort()`
+  (#2469, #2544)
+[core] [012] Added `JobConfig.suspendOnFailure`: suspend a job on exception instead of
+  cancelling it (#2411)
+[cdc] Improved the consistency of reconnect behaviour across CDC sources, new uniform API
+  to configure the reconnect strategy (#2419)
+
+2. Enhancements
+
+[core] @guenter-hesse contributed a test source to benchmark Jet's throughput and latency (#2382)
+[core] [013] Improved watermark semantics that prevent low event rate from stalling an
+  ingestion time-based pipeline (#2485, #2514)
+[cdc] Exposed the sequence number in the CDC `ChangeRecord` that orders the events (#2390)
+[core] Two new DAG edge types: `distributeToOne` (sending all data to one member) and `ordered`
+  (maintaining the sort order) (#2394, #2469, #2544)
+[core] Disabled access to external XML entities when parsing XML config, this was a potential
+  XXE attack vector (#2528)
+
+3. Fixes
+
+[core] Fixed error handling during job startup that could result in inconsistent job state (#2383)
+[core] Fixed an internal exception that leaked out of Observable (#2313, #2389)
+[core] Prevented Observable from processing in-flight items after cancellation (#2415, #2418)
+[cli] @caioguedes fixed an issue with `--targets` option in CLI where it would overwrite other
+  settings (#2373, #2421)
+[metrics] Fixed a problem where an internally added DAG vertex would show up as a source instead
+  of the actual source vertex (#2475, #2476)
+[core] Fixed a race that could cause `getJobStatus()` to throw an exception if called right after
+  `newJob()` (#2481, #2484)
+[core] Fix a race between snapshotting and restarting (#2487, #2503)
+[core] Fixed a race where `getJobStatus()` would report `RUNNING` even though it was actually
+  `COMPLETING`. (#2507)
+[core] Fixed an issue where a `DONE_ITEM` could get lost due to connection failure, preventing
+  the job from completing (#2158, #2532)
+[core] Fixed a job failure related to the coordinator member failing (#2461, #2546)
+[core] Fixed a job failure related to a member reconnecting (#2542, #2547)
+[core] Improved robustness related to Jet's internal `IMap` operations (#2533, #2550)
+[cdc] Upgraded Jackson jr dep, solving a null handling issue (#2459)
+[cdc] Upgraded the Debezium dep, solving a Postgress issue resulting in data loss when
+  snapshotting (#2406)
+[core] Fixed a bug where a non-keyed aggregating stage would produce no output when no input
+  (#2560, #2567)
+
+4. Breaking Changes
+
+[pipeline-api] Breaking signature change to `Sources.streamFromProcessorWithWatermarks()`
+[pipeline-api] Deprecated `Pipeline.toDag()`, made `Pipeline` and all its components `Serializable`.
+[core-api] Breaking signature change to `StreamEventJournalP`, methods `streamRemoteMapSupplier()`
+  and `streamRemoteCacheSupplier`
+
+
 ** Hazelcast Jet 4.2 **
 
 Members of the open source community that appear in these release notes:
@@ -70,6 +134,7 @@ Thank you for your valuable contributions!
 [jet-cli] Deprecated --address and --cluster-name options from command-line as --targets
   should now be used instead. (#2276)
 
+
 ** Hazelcast Jet 4.1.1 **
 
 1. Enhancements