lincoln-lil commented on code in PR #721:
URL: https://github.com/apache/flink-web/pull/721#discussion_r1522831889
##########
docs/content/posts/2024-03-xx-release-1.19.0.md:
##########
@@ -0,0 +1,501 @@
+---
+authors:
+- LincolnLee:
+ name: "Lincoln Lee"
+ twitter: lincoln_86xy
+
+date: "2024-03-xxT22:00:00Z"
+subtitle: ""
+title: Announcing the Release of Apache Flink 1.19
+aliases:
+- /news/2024/03/xx/release-1.19.0.html
+---
+
+The Apache Flink PMC is pleased to announce the release of Apache Flink
1.19.0. As usual, we are
+looking at a packed release with a wide variety of improvements and new
features. Overall, 162
+people contributed to this release completing 33 FLIPs and 600+ issues. Thank
you!
+
+Let's dive into the highlights.
+
+# Flink SQL Improvements
+
+## Custom Parallelism for Table/SQL Sources
+
+Now in Flink 1.19, you can set a custom parallelism for performance tuning via
the `scan.parallelism`
+option. The first available connector is DataGen (Kafka connector is on the
way). Here is an example
+using SQL Client:
+
+```sql
+-- set parallelism within the ddl
+CREATE TABLE Orders (
+ order_number BIGINT,
+ price DECIMAL(32,2),
+ buyer ROW<first_name STRING, last_name STRING>,
+ order_time TIMESTAMP(3)
+) WITH (
+ 'connector' = 'datagen',
+ 'scan.parallelism' = '4'
+);
+
+-- or set parallelism via dynamic table option
+SELECT * FROM Orders /*+ OPTIONS('scan.parallelism'='4') */;
+```
+
+**More Information**
+*
[Documentation](https://nightlies.apache.org/flink/flink-docs-release-1.19/docs/dev/table/sourcessinks/#scan-table-source)
+* [FLIP-367: Support Setting Parallelism for Table/SQL
Sources](https://cwiki.apache.org/confluence/pages/viewpage.action?pageId=263429150)
+
+
+## Configurable SQL Gateway Java Options
+
+A new option `env.java.opts.sql-gateway` for specifying the Java options is
introduced in Flink 1.19,
+so you can fine-tune the memory settings, garbage collection behavior, and
other relevant Java
+parameters for SQL Gateway.
+
+**More Information**
+* [FLINK-33203](https://issues.apache.org/jira/browse/FLINK-33203)
+
+
+## Configure Different State TTLs Using SQL Hint
+
+Starting from Flink 1.18, Table API and SQL users can set state time-to-live
(TTL) individually for
+stateful operators via the SQL compiled plan. In Flink 1.19, users have a more
flexible way to
+specify custom TTL values for regular joins and group aggregations directly
within their queries by [utilizing the STATE_TTL
hint](https://nightlies.apache.org/flink/flink-docs-release-1.19/docs/dev/table/sql/queries/hints/#state-ttl-hints).
+This improvement means that you no longer need to alter your compiled plan to
set specific TTLs for
+these frequently used operators. With the introduction of `STATE_TTL` hints,
you can streamline your workflow and
+dynamically adjust the TTL based on your operational requirements.
+
+Here is an example:
+```sql
+-- set state ttl for join
+SELECT /*+ STATE_TTL('Orders'= '1d', 'Customers' = '20d') */ *
+FROM Orders LEFT OUTER JOIN Customers
+ ON Orders.o_custkey = Customers.c_custkey;
+
+-- set state ttl for aggregation
+SELECT /*+ STATE_TTL('o' = '1d') */ o_orderkey, SUM(o_totalprice) AS revenue
+FROM Orders AS o
+GROUP BY o_orderkey;
+```
+
+**More Information**
+*
[Documentation](https://nightlies.apache.org/flink/flink-docs-release-1.19/docs/dev/table/sql/queries/hints/#state-ttl-hints)
+* [FLIP-373: Support Configuring Different State TTLs using SQL
Hint](https://cwiki.apache.org/confluence/display/FLINK/FLIP-373%3A+Support+Configuring+Different+State+TTLs+using+SQL+Hint)
+
+
+## Named Parameters for Functions and Procedures
+
+Named parameters now can be used when calling a function or stored procedure.
With named parameters,
+users do not need to strictly specify the parameter position, just specify the
parameter name and its
+corresponding value. At the same time, if non-essential parameters are not
specified, they will default to being filled with null.
+
+Here's an example of defining a function with one mandatory parameter and two
optional parameters using named parameters:
+```java
+public static class NamedArgumentsTableFunction extends TableFunction<Object> {
+
+ @FunctionHint(
+ output = @DataTypeHint("STRING"),
+ arguments = {
+ @ArgumentHint(name = "in1", isOptional
= false, type = @DataTypeHint("STRING")),
+ @ArgumentHint(name = "in2", isOptional
= true, type = @DataTypeHint("STRING")),
+ @ArgumentHint(name = "in3", isOptional
= true, type = @DataTypeHint("STRING"))})
+ public void eval(String arg1, String arg2, String arg3) {
+ collect(arg1 + ", " + arg2 + "," + arg3);
+ }
+
+}
+```
+When calling the function in SQL, parameters can be specified by name, for
example:
+```sql
+SELECT * FROM TABLE(myFunction(in1 => 'v1', in3 => 'v3', in2 => 'v2'));
+```
+Also the optional parameters can be omitted:
+```sql
+SELECT * FROM TABLE(myFunction(in1 => 'v1'));
+```
+
+**More Information**
+*
[Documentation](https://nightlies.apache.org/flink/flink-docs-release-1.19/docs/dev/table/functions/udfs/#named-parameters)
+* [FLIP-387: Support named parameters for functions and call
procedures](https://cwiki.apache.org/confluence/display/FLINK/FLIP-387%3A+Support+named+parameters+for+functions+and+call+procedures)
+
+## Window TVF Aggregation Features
+
+* Supports SESSION Window TVF in Streaming Mode<br />
+Now users can use SESSION Window TVF in streaming mode. A simple example is as
follows:
+```sql
+-- session window with partition keys
+SELECT * FROM TABLE(
+ SESSION(TABLE Bid PARTITION BY item, DESCRIPTOR(bidtime), INTERVAL '5'
MINUTES));
+
+-- apply aggregation on the session windowed table with partition keys
+SELECT window_start, window_end, item, SUM(price) AS total_price
+FROM TABLE(
+ SESSION(TABLE Bid PARTITION BY item, DESCRIPTOR(bidtime), INTERVAL '5'
MINUTES))
+GROUP BY item, window_start, window_end;
+```
+* Supports Changelog Inputs for Window TVF Aggregation<br />
+ Window aggregation operators (generated based on Window TVF Function) can
now handle changelog
+ streams (e.g., CDC data sources, etc.). Users are recommended to migrate
from legacy window
+ aggregation to the new syntax for more complete feature support.
+
+**More Information**
+*
[Documentation](https://nightlies.apache.org/flink/flink-docs-release-1.19/docs/dev/table/sql/queries/window-tvf/#session)
+
+## New UDF Type: AsyncScalarFunction
+
+The common UDF type `ScalarFunction` works well for CPU-intensive operations,
but less well for IO
+bound or otherwise long-running computations. In Flink 1.19, we have a new
`AsyncScalarFunction`
+which is a user-defined asynchronous `ScalarFunction` allows for issuing
concurrent function calls
+asynchronously.
+
+**More Information**
+* [FLIP-400: AsyncScalarFunction for asynchronous scalar function
support](https://cwiki.apache.org/confluence/display/FLINK/FLIP-400%3A+AsyncScalarFunction+for+asynchronous+scalar+function+support)
+
+## Tuning: MiniBatch Optimization for Regular Joins
+
+The record amplification is a pain point when performing cascading joins in
Flink, now in Flink 1.19,
+the new mini-batch optimization can be used for regular join to reduce
intermediate result in such
+cascading join scenarios.
+
+<div style="text-align: center;">
+<img src="/img/blog/2024-03-xx-release-1.19.0/minibatch_join.png"
style="width:90%;margin:15px">
+</div>
+
+**More Information**
+*
[minibatch-regular-joins](https://nightlies.apache.org/flink/flink-docs-release-1.19/docs/dev/table/tuning/#minibatch-regular-joins)
+* [FLIP-415: Introduce a new join operator to support
minibatch](https://cwiki.apache.org/confluence/display/FLINK/FLIP-415%3A+Introduce+a+new+join+operator+to+support+minibatch)
+
+# Runtime & Coordination Improvements
+
+## Dynamic Source Parallelism Inference for Batch Jobs
+
+In Flink 1.19, we have supported dynamic source parallelism inference for
batch jobs, which allows
+source connectors to dynamically infer the parallelism based on the actual
amount of data to consume.
+This feature is a significant improvement over previous versions, which only
assigned a fixed default
+parallelism to source vertices.
+Source connectors need to implement the inference interface to enable dynamic
parallelism inference.
+Currently, the FileSource connector has already been developed with this
functionality in place.
+Additionally, the configuration
`execution.batch.adaptive.auto-parallelism.default-source-parallelism`
+will be used as the upper bound of source parallelism inference. And now it
will not default to 1.
+Instead, if it is not set, the upper bound of allowed parallelism set via
+`execution.batch.adaptive.auto-parallelism.max-parallelism` will be used. If
that configuration is
+also not set, the default parallelism set via `parallelism.default` or
`StreamExecutionEnvironment#setParallelism()`
+will be used instead.
+
+**More Information**
+*
[Documentation](https://nightlies.apache.org/flink/flink-docs-release-1.19/docs/deployment/elastic_scaling/#enable-dynamic-parallelism-inference-support-for-sources)
+* [FLIP-379: Support dynamic source parallelism inference for batch
jobs](https://cwiki.apache.org/confluence/display/FLINK/FLIP-379%3A+Dynamic+source+parallelism+inference+for+batch+jobs)
+
+## Standard YAML for Flink Configuration
+
+Starting with Flink 1.19, Flink has officially introduced full support for the
standard YAML 1.2
+syntax. The default configuration file has been changed to `config.yaml` and
placed in the `conf/`
+directory. Users should directly modify this file to configure Flink.
+If users want to use the legacy configuration file `flink-conf.yaml`, users
just need to copy this
+file into the `conf/` directory. Once the legacy configuration file
`flink-conf.yaml` is detected,
+Flink will prioritize using it as the configuration file. And in the upcoming
Flink 2.0, the
+`flink-conf.yaml` configuration file will no longer work.
+
+**More Information**
+*
[flink-configuration-file](https://nightlies.apache.org/flink/flink-docs-release-1.19/docs/deployment/config/#flink-configuration-file)
+* [FLIP-366: Support standard YAML for Flink
configuration](https://cwiki.apache.org/confluence/display/FLINK/FLIP-366%3A+Support+standard+YAML+for+FLINK+configuration?src=contextnavpagetreemode)
+
+## Profiling JobManager/TaskManager on Flink Web
+
+In Flink 1.19, we support triggering profiling at the JobManager/TaskManager
level, allowing users to
+create a profiling instance with arbitrary intervals and event modes
(supported by
[async-profiler](https://github.com/async-profiler/async-profiler)).
+Users can easily submit profiles and export results in the Flink Web UI.
+
+For example, users can simply submit a profiling instance with a specified
period and mode by
+"Creating a Profiling Instance" after identifying a candidate
TaskManager/JobManager with a
+performance bottleneck:
+<div style="text-align: center;">
+<img src="/img/blog/2024-03-xx-release-1.19.0/profiling.png"
style="width:90%;margin:15px">
+</div>
+
+then easily download the interactive HTML file after the profiling instance is
complete:
+<div style="text-align: center;">
+<img src="/img/blog/2024-03-xx-release-1.19.0/profiling-res.png"
style="width:90%;margin:15px">
+</div>
+
+**More Information**
+*
[Documentation](https://nightlies.apache.org/flink/flink-docs-release-1.19/docs/ops/debugging/profiler/)
+* [FLIP-375: Built-in cross-platform powerful java
profiler](https://cwiki.apache.org/confluence/x/64lEE)
+
+## New Config Options for Administrator JVM Options
+
+A set of administrator JVM options are available, which prepend the user-set
extra JVM options for
+platform-wide JVM tuning.
+
+**More Information**
+*
[Documentation](https://nightlies.apache.org/flink/flink-docs-release-1.19/docs/deployment/config/#jvm-and-logging-options)
+* [FLIP-397: Add config options for administrator JVM
options](https://cwiki.apache.org/confluence/display/FLINK/FLIP-397%3A+Add+config+options+for+administrator+JVM+options?src=jira)
+
+## Beta Support for Java 21
+
+Apache Flink was made ready to compile and run with Java 21. This feature is
still in beta mode.
+Issues should be reported in Flink's bug tracker.
+
+**More Information**
+* [FLINK-33163](https://issues.apache.org/jira/browse/FLINK-33163)
+
+# Checkpoints Improvements
+
+## Using Larger Checkpointing Interval When Source is Processing Backlog
+
+`IsProcessingBacklog` is introduced to indicate whether a record should be
processed with low latency
+or high throughput. Connector developers can update the Source implementation
to utilize the
+`SplitEnumeratorContext#setIsProcessingBacklog` method to report whether the
records are backlog records.
+Users can set the `execution.checkpointing.interval-during-backlog` to use a
larger checkpoint interval
+to enhance the throughput while the job is processing backlog if the source is
backlog-aware.
+
+**More Information**
+* [FLINK-32514](https://issues.apache.org/jira/browse/FLINK-32514)
+* [FLIP-309: Support using larger checkpointing interval when source is
processing
backlog](https://cwiki.apache.org/confluence/display/FLINK/FLIP-309%3A+Support+using+larger+checkpointing+interval+when+source+is+processing+backlog)
+
+## CheckpointsCleaner Clean Individual Checkpoint States in Parallel
+
+Now when disposing of no longer needed checkpoints, every state handle/state
file will be disposed
+in parallel by the ioExecutor, vastly improving the disposing speed of a
single checkpoint (for
+large checkpoints the disposal time can be improved from 10 minutes to < 1
minute) . The old
+behavior can be restored by setting `state.checkpoint.cleaner.parallel-mode`
to false.
+
+**More Information**
+* [FLINK-33090](https://issues.apache.org/jira/browse/FLINK-33090)
+
+## Trigger Checkpoints through Command Line Client
+
+The command line interface supports triggering a checkpoint manually. Usage:
+```shell
+./bin/flink checkpoint $JOB_ID [-full]
+```
+By specifying the '-full' option, a full checkpoint is triggered. Otherwise an
incremental
+checkpoint is triggered if the job is configured to take incremental ones
periodically.
+
+**More Information**
+* [FLINK-6755](https://issues.apache.org/jira/browse/FLINK-6755)
+
+# Connector API Improvements
+
+## New Interfaces to SinkV2 That Are Consistent with SourceV2
Review Comment:
@PatrickRen Thanks for the correction! I directly use the title of
FLINK-33973, but it's true that we don't have a class like `Sourcev2` in the
source code, IIUC, "v2 " source should actually refer to FLIP-27 source.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscr...@flink.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
