This is an automated email from the ASF dual-hosted git repository.
kurt pushed a commit to branch release-1.11
in repository https://gitbox.apache.org/repos/asf/flink.git
The following commit(s) were added to refs/heads/release-1.11 by this push:
new a450571 [FLINK-17340][docs] Update docs which related to default
planner changing.
a450571 is described below
commit a450571c354832fe792d324beddbcc6e98cafb09
Author: Kurt Young <k...@apache.org>
AuthorDate: Tue Jun 2 09:50:23 2020 +0800
[FLINK-17340][docs] Update docs which related to default planner changing.
This closes #12429
(cherry picked from commit 1c78ab397de524836fd69c6218b1122aa387c251)
---
docs/dev/table/catalogs.md | 4 ++--
docs/dev/table/catalogs.zh.md | 4 ++--
docs/dev/table/common.md | 52 ++++++++++++++++++-----------------------
docs/dev/table/common.zh.md | 51 ++++++++++++++++++----------------------
docs/dev/table/hive/index.md | 4 ++--
docs/dev/table/hive/index.zh.md | 4 ++--
docs/dev/table/index.md | 6 ++---
docs/dev/table/index.zh.md | 6 ++---
8 files changed, 60 insertions(+), 71 deletions(-)
diff --git a/docs/dev/table/catalogs.md b/docs/dev/table/catalogs.md
index 961d0a6..7236154 100644
--- a/docs/dev/table/catalogs.md
+++ b/docs/dev/table/catalogs.md
@@ -63,7 +63,7 @@ Set a `JdbcCatalog` with the following parameters:
<div data-lang="Java" markdown="1">
{% highlight java %}
-EnvironmentSettings settings =
EnvironmentSettings.newInstance().useBlinkPlanner().inStreamingMode().build();
+EnvironmentSettings settings =
EnvironmentSettings.newInstance().inStreamingMode().build();
TableEnvironment tableEnv = TableEnvironment.create(settings);
String name = "mypg";
@@ -82,7 +82,7 @@ tableEnv.useCatalog("mypg");
<div data-lang="Scala" markdown="1">
{% highlight scala %}
-val settings =
EnvironmentSettings.newInstance().useBlinkPlanner().inStreamingMode().build()
+val settings = EnvironmentSettings.newInstance().inStreamingMode().build()
val tableEnv = TableEnvironment.create(settings)
val name = "mypg"
diff --git a/docs/dev/table/catalogs.zh.md b/docs/dev/table/catalogs.zh.md
index 23e0969..f775ad6 100644
--- a/docs/dev/table/catalogs.zh.md
+++ b/docs/dev/table/catalogs.zh.md
@@ -59,7 +59,7 @@ Set a `Jdbcatalog` with the following parameters:
<div data-lang="Java" markdown="1">
{% highlight java %}
-EnvironmentSettings settings =
EnvironmentSettings.newInstance().useBlinkPlanner().inStreamingMode().build();
+EnvironmentSettings settings =
EnvironmentSettings.newInstance().inStreamingMode().build();
TableEnvironment tableEnv = TableEnvironment.create(settings);
String name = "mypg";
@@ -78,7 +78,7 @@ tableEnv.useCatalog("mypg");
<div data-lang="Scala" markdown="1">
{% highlight scala %}
-val settings =
EnvironmentSettings.newInstance().useBlinkPlanner().inStreamingMode().build()
+val settings = EnvironmentSettings.newInstance().inStreamingMode().build()
val tableEnv = TableEnvironment.create(settings)
val name = "mypg"
diff --git a/docs/dev/table/common.md b/docs/dev/table/common.md
index 4dae5b7..a6fe84f 100644
--- a/docs/dev/table/common.md
+++ b/docs/dev/table/common.md
@@ -32,12 +32,11 @@ Main Differences Between the Two Planners
1. Blink treats batch jobs as a special case of streaming. As such, the
conversion between Table and DataSet is also not supported, and batch jobs will
not be translated into `DateSet` programs but translated into `DataStream`
programs, the same as the streaming jobs.
2. The Blink planner does not support `BatchTableSource`, use bounded
`StreamTableSource` instead of it.
-3. The Blink planner only support the brand new `Catalog` and does not support
`ExternalCatalog` which is deprecated.
-4. The implementations of `FilterableTableSource` for the old planner and the
Blink planner are incompatible. The old planner will push down
`PlannerExpression`s into `FilterableTableSource`, while the Blink planner will
push down `Expression`s.
-5. String based key-value config options (Please see the documentation about
[Configuration]({{ site.baseurl }}/dev/table/config.html) for details) are only
used for the Blink planner.
-6. The implementation(`CalciteConfig`) of `PlannerConfig` in two planners is
different.
-7. The Blink planner will optimize multiple-sinks into one DAG (supported only
on `TableEnvironment`, not on `StreamTableEnvironment`). The old planner will
always optimize each sink into a new DAG, where all DAGs are independent of
each other.
-8. The old planner does not support catalog statistics now, while the Blink
planner does.
+3. The implementations of `FilterableTableSource` for the old planner and the
Blink planner are incompatible. The old planner will push down
`PlannerExpression`s into `FilterableTableSource`, while the Blink planner will
push down `Expression`s.
+4. String based key-value config options (Please see the documentation about
[Configuration]({{ site.baseurl }}/dev/table/config.html) for details) are only
used for the Blink planner.
+5. The implementation(`CalciteConfig`) of `PlannerConfig` in two planners is
different.
+6. The Blink planner will optimize multiple-sinks into one DAG (supported only
on `TableEnvironment`, not on `StreamTableEnvironment`). The old planner will
always optimize each sink into a new DAG, where all DAGs are independent of
each other.
+7. The old planner does not support catalog statistics now, while the Blink
planner does.
Structure of Table API and SQL Programs
@@ -831,6 +830,19 @@ Translate and Execute a Query
The behavior of translating and executing a query is different for the two
planners.
<div class="codetabs" markdown="1">
+
+<div data-lang="Blink planner" markdown="1">
+Table API and SQL queries are translated into [DataStream]({{ site.baseurl
}}/dev/datastream_api.html) programs whether their input is streaming or batch.
A query is internally represented as a logical query plan and is translated in
two phases:
+
+1. Optimization of the logical plan,
+2. Translation into a DataStream program.
+
+a Table API or SQL query is translated when:
+
+* `TableEnvironment.execute()` is called. A `Table` (emitted to a `TableSink`
through `Table.insertInto()`) or a SQL update query (specified through
`TableEnvironment.sqlUpdate()`) will be buffered in `TableEnvironment` first.
All sinks will be optimized into one DAG.
+* A `Table` is translated when it is converted into a `DataStream` (see
[Integration with DataStream and DataSet
API](#integration-with-datastream-and-dataset-api)). Once translated, it's a
regular DataStream program and is executed when
`StreamExecutionEnvironment.execute()` is called.
+</div>
+
<div data-lang="Old planner" markdown="1">
Table API and SQL queries are translated into [DataStream]({{ site.baseurl
}}/dev/datastream_api.html) or [DataSet]({{ site.baseurl }}/dev/batch) programs
depending on whether their input is a streaming or batch input. A query is
internally represented as a logical query plan and is translated in two phases:
@@ -849,22 +861,8 @@ For batch, a Table API or SQL query is translated when:
* a `Table` is converted into a `DataSet` (see [Integration with DataStream
and DataSet API](#integration-with-datastream-and-dataset-api)).
Once translated, a Table API or SQL query is handled like a regular DataSet
program and is executed when `ExecutionEnvironment.execute()` is called.
-
</div>
-<div data-lang="Blink planner" markdown="1">
-Table API and SQL queries are translated into [DataStream]({{ site.baseurl
}}/dev/datastream_api.html) programs whether their input is streaming or batch.
A query is internally represented as a logical query plan and is translated in
two phases:
-
-1. Optimization of the logical plan,
-2. Translation into a DataStream program.
-
-a Table API or SQL query is translated when:
-
-* `TableEnvironment.execute()` is called. A `Table` (emitted to a `TableSink`
through `Table.insertInto()`) or a SQL update query (specified through
`TableEnvironment.sqlUpdate()`) will be buffered in `TableEnvironment` first.
All sinks will be optimized into one DAG.
-* A `Table` is translated when it is converted into a `DataStream` (see
[Integration with DataStream and DataSet
API](#integration-with-datastream-and-dataset-api)). Once translated, it's a
regular DataStream program and is executed when
`StreamExecutionEnvironment.execute()` is called.
-
-
-</div>
</div>
{% top %}
@@ -1407,16 +1405,7 @@ Query Optimization
------------------
<div class="codetabs" markdown="1">
-<div data-lang="Old planner" markdown="1">
-
-Apache Flink leverages Apache Calcite to optimize and translate queries. The
optimization currently performed include projection and filter push-down,
subquery decorrelation, and other kinds of query rewriting. Old planner does
not yet optimize the order of joins, but executes them in the same order as
defined in the query (order of Tables in the `FROM` clause and/or order of join
predicates in the `WHERE` clause).
-
-It is possible to tweak the set of optimization rules which are applied in
different phases by providing a `CalciteConfig` object. This can be created via
a builder by calling `CalciteConfig.createBuilder())` and is provided to the
TableEnvironment by calling
`tableEnv.getConfig.setPlannerConfig(calciteConfig)`.
-
-</div>
-
<div data-lang="Blink planner" markdown="1">
-
Apache Flink leverages and extends Apache Calcite to perform sophisticated
query optimization.
This includes a series of rule and cost-based optimizations such as:
@@ -1436,7 +1425,12 @@ This includes a series of rule and cost-based
optimizations such as:
The optimizer makes intelligent decisions, based not only on the plan but also
rich statistics available from the data sources and fine-grain costs for each
operator such as io, cpu, network, and memory.
Advanced users may provide custom optimizations via a `CalciteConfig` object
that can be provided to the table environment by calling
`TableEnvironment#getConfig#setPlannerConfig`.
+</div>
+<div data-lang="Old planner" markdown="1">
+Apache Flink leverages Apache Calcite to optimize and translate queries. The
optimization currently performed include projection and filter push-down,
subquery decorrelation, and other kinds of query rewriting. Old planner does
not yet optimize the order of joins, but executes them in the same order as
defined in the query (order of Tables in the `FROM` clause and/or order of join
predicates in the `WHERE` clause).
+
+It is possible to tweak the set of optimization rules which are applied in
different phases by providing a `CalciteConfig` object. This can be created via
a builder by calling `CalciteConfig.createBuilder())` and is provided to the
TableEnvironment by calling
`tableEnv.getConfig.setPlannerConfig(calciteConfig)`.
</div>
</div>
diff --git a/docs/dev/table/common.zh.md b/docs/dev/table/common.zh.md
index 08fef48..7bb7954 100644
--- a/docs/dev/table/common.zh.md
+++ b/docs/dev/table/common.zh.md
@@ -32,12 +32,11 @@ Table API 和 SQL 集成在同一套 API 中。这套 API 的核心概念是`Tab
1. Blink 将批处理作业视作流处理的一种特例。严格来说，`Table` 和 `DataSet` 之间不支持相互转换，并且批处理作业也不会转换成
`DataSet` 程序而是转换成 `DataStream` 程序，流处理作业也一样。
2. Blink 计划器不支持 `BatchTableSource`，而是使用有界的 `StreamTableSource` 来替代。
-3. Blink 计划器仅支持全新的 `Catalog` 不支持被弃用的 `ExternalCatalog`。
-4. 旧计划器和 Blink 计划器中 `FilterableTableSource` 的实现是不兼容的。旧计划器会将
`PlannerExpression` 下推至 `FilterableTableSource`，而 Blink 计划器则是将 `Expression` 下推。
-5. 基于字符串的键值配置选项仅在 Blink 计划器中使用。（详情参见 [配置]({{ site.baseurl
}}/zh/dev/table/config.html) ）
-6. `PlannerConfig` 在两种计划器中的实现（`CalciteConfig`）是不同的。
-7. Blink 计划器会将多sink（multiple-sinks）优化成一张有向无环图（DAG）（仅支持 `TableEnvironment`，不支持
`StreamTableEnvironment`）。旧计划器总是将每个sink都优化成一个新的有向无环图，且所有图相互独立。
-8. 旧计划器目前不支持 catalog 统计数据，而 Blink 支持。
+3. 旧计划器和 Blink 计划器中 `FilterableTableSource` 的实现是不兼容的。旧计划器会将
`PlannerExpression` 下推至 `FilterableTableSource`，而 Blink 计划器则是将 `Expression` 下推。
+4. 基于字符串的键值配置选项仅在 Blink 计划器中使用。（详情参见 [配置]({{ site.baseurl
}}/zh/dev/table/config.html) ）
+5. `PlannerConfig` 在两种计划器中的实现（`CalciteConfig`）是不同的。
+6. Blink 计划器会将多sink（multiple-sinks）优化成一张有向无环图（DAG）（仅支持 `TableEnvironment`，不支持
`StreamTableEnvironment`）。旧计划器总是将每个sink都优化成一个新的有向无环图，且所有图相互独立。
+7. 旧计划器目前不支持 catalog 统计数据，而 Blink 支持。
Table API 和 SQL 程序的结构
@@ -810,6 +809,19 @@ result.insert_into("CsvSinkTable")
两种计划器翻译和执行查询的方式是不同的。
<div class="codetabs" markdown="1">
+
+<div data-lang="Blink planner" markdown="1">
+不论输入数据源是流式的还是批式的，Table API 和 SQL 查询都会被转换成 [DataStream]({{ site.baseurl
}}/zh/dev/datastream_api.html) 程序。查询在内部表示为逻辑查询计划，并被翻译成两个阶段：
+
+1. 优化逻辑执行计划
+2. 翻译成 DataStream 程序
+
+Table API 或者 SQL 查询在下列情况下会被翻译：
+
+* 当 `TableEnvironment.execute()` 被调用时。`Table` （通过 `Table.insertInto()` 输出给
`TableSink`）和 SQL （通过调用 `TableEnvironment.sqlUpdate()`）会先被缓存到
`TableEnvironment` 中，所有的 sink 会被优化成一张有向无环图。
+* `Table` 被转换成 `DataStream` 时（参阅[与 DataStream 和 DataSet API
结合](#integration-with-datastream-and-dataset-api)）。转换完成后，它就成为一个普通的 DataStream
程序，并且会在调用 `StreamExecutionEnvironment.execute()` 的时候被执行。
+</div>
+
<div data-lang="Old planner" markdown="1">
Table API 和 SQL 查询会被翻译成 [DataStream]({{ site.baseurl
}}/zh/dev/datastream_api.html) 或者 [DataSet]({{ site.baseurl }}/zh/dev/batch)
程序， 这取决于它们的输入数据源是流式的还是批式的。查询在内部表示为逻辑查询计划，并被翻译成两个阶段：
@@ -828,21 +840,8 @@ Table API 和 SQL 查询会被翻译成 [DataStream]({{ site.baseurl
}}/zh/dev/d
* `Table` 被转换成 `DataSet` 时（参阅[与 DataStream 和 DataSet API
结合](#integration-with-datastream-and-dataset-api)）。
翻译完成后，Table API 或者 SQL 查询会被当做普通的 DataSet 程序对待并且会在调用
`ExecutionEnvironment.execute()` 的时候被执行。
-
</div>
-<div data-lang="Blink planner" markdown="1">
-不论输入数据源是流式的还是批式的，Table API 和 SQL 查询都会被转换成 [DataStream]({{ site.baseurl
}}/zh/dev/datastream_api.html) 程序。查询在内部表示为逻辑查询计划，并被翻译成两个阶段：
-
-1. 优化逻辑执行计划
-2. 翻译成 DataStream 程序
-
-Table API 或者 SQL 查询在下列情况下会被翻译：
-
-* 当 `TableEnvironment.execute()` 被调用时。`Table` （通过 `Table.insertInto()` 输出给
`TableSink`）和 SQL （通过调用 `TableEnvironment.sqlUpdate()`）会先被缓存到
`TableEnvironment` 中，所有的 sink 会被优化成一张有向无环图。
-* `Table` 被转换成 `DataStream` 时（参阅[与 DataStream 和 DataSet API
结合](#integration-with-datastream-and-dataset-api)）。转换完成后，它就成为一个普通的 DataStream
程序，并且会在调用 `StreamExecutionEnvironment.execute()` 的时候被执行。
-
-</div>
</div>
{% top %}
@@ -1388,16 +1387,7 @@ val table: Table = tableEnv.fromDataStream(stream,
$"name" as "myName")
------------------
<div class="codetabs" markdown="1">
-<div data-lang="Old planner" markdown="1">
-
-Apache Flink 利用 Apache Calcite
来优化和翻译查询。当前执行的优化包括投影和过滤器下推，子查询消除以及其他类型的查询重写。原版计划程序尚未优化 join
的顺序，而是按照查询中定义的顺序执行它们（FROM 子句中的表顺序和/或 WHERE 子句中的 join 谓词顺序）。
-
-通过提供一个 `CalciteConfig` 对象，可以调整在不同阶段应用的优化规则集合。这个对象可以通过调用构造器
`CalciteConfig.createBuilder()` 创建，并通过调用
`tableEnv.getConfig.setPlannerConfig(calciteConfig)` 提供给 TableEnvironment。
-
-</div>
-
<div data-lang="Blink planner" markdown="1">
-
Apache Flink 使用并扩展了 Apache Calcite 来执行复杂的查询优化。
这包括一系列基于规则和成本的优化，例如：
@@ -1417,7 +1407,12 @@ Apache Flink 使用并扩展了 Apache Calcite 来执行复杂的查询优化。
优化器不仅基于计划，而且还基于可从数据源获得的丰富统计信息以及每个算子（例如 io，cpu，网络和内存）的细粒度成本来做出明智的决策。
高级用户可以通过 `CalciteConfig` 对象提供自定义优化，可以通过调用
`TableEnvironment＃getConfig＃setPlannerConfig` 将其提供给 TableEnvironment。
+</div>
+<div data-lang="Old planner" markdown="1">
+Apache Flink 利用 Apache Calcite
来优化和翻译查询。当前执行的优化包括投影和过滤器下推，子查询消除以及其他类型的查询重写。原版计划程序尚未优化 join
的顺序，而是按照查询中定义的顺序执行它们（FROM 子句中的表顺序和/或 WHERE 子句中的 join 谓词顺序）。
+
+通过提供一个 `CalciteConfig` 对象，可以调整在不同阶段应用的优化规则集合。这个对象可以通过调用构造器
`CalciteConfig.createBuilder()` 创建，并通过调用
`tableEnv.getConfig.setPlannerConfig(calciteConfig)` 提供给 TableEnvironment。
</div>
</div>
diff --git a/docs/dev/table/hive/index.md b/docs/dev/table/hive/index.md
index 3698df2..2bfa55b 100644
--- a/docs/dev/table/hive/index.md
+++ b/docs/dev/table/hive/index.md
@@ -311,7 +311,7 @@ Take Hive version 2.3.4 for example:
<div data-lang="Java" markdown="1">
{% highlight java %}
-EnvironmentSettings settings =
EnvironmentSettings.newInstance().useBlinkPlanner().inBatchMode().build();
+EnvironmentSettings settings =
EnvironmentSettings.newInstance().inBatchMode().build();
TableEnvironment tableEnv = TableEnvironment.create(settings);
String name = "myhive";
@@ -329,7 +329,7 @@ tableEnv.useCatalog("myhive");
<div data-lang="Scala" markdown="1">
{% highlight scala %}
-val settings =
EnvironmentSettings.newInstance().useBlinkPlanner().inBatchMode().build()
+val settings = EnvironmentSettings.newInstance().inBatchMode().build()
val tableEnv = TableEnvironment.create(settings)
val name = "myhive"
diff --git a/docs/dev/table/hive/index.zh.md b/docs/dev/table/hive/index.zh.md
index 8136a3b..fae835e 100644
--- a/docs/dev/table/hive/index.zh.md
+++ b/docs/dev/table/hive/index.zh.md
@@ -307,7 +307,7 @@ Apache Hive 是基于 Hadoop 之上构建的, 首先您需要 Hadoop 的依赖
<div data-lang="Java" markdown="1">
{% highlight java %}
-EnvironmentSettings settings =
EnvironmentSettings.newInstance().useBlinkPlanner().inBatchMode().build();
+EnvironmentSettings settings =
EnvironmentSettings.newInstance().inBatchMode().build();
TableEnvironment tableEnv = TableEnvironment.create(settings);
String name = "myhive";
@@ -325,7 +325,7 @@ tableEnv.useCatalog("myhive");
<div data-lang="Scala" markdown="1">
{% highlight scala %}
-val settings =
EnvironmentSettings.newInstance().useBlinkPlanner().inBatchMode().build()
+val settings = EnvironmentSettings.newInstance().inBatchMode().build()
val tableEnv = TableEnvironment.create(settings)
val name = "myhive"
diff --git a/docs/dev/table/index.md b/docs/dev/table/index.md
index 165f7de..eb6b700 100644
--- a/docs/dev/table/index.md
+++ b/docs/dev/table/index.md
@@ -38,7 +38,7 @@ Starting from Flink 1.9, Flink provides two different planner
implementations fo
translating relational operators into an executable, optimized Flink job. Both
of the planners come with different optimization rules and runtime classes.
They may also differ in the set of supported features.
-<span class="label label-danger">Attention</span> For production use cases, we
recommend the old planner that was present before Flink 1.9 for now.
+<span class="label label-danger">Attention</span> For production use cases, we
recommend the blink planner that has become the default planner since 1.11.
All Table API and SQL components are bundled in the `flink-table` or
`flink-table-blink` Maven artifacts.
@@ -49,8 +49,8 @@ The following dependencies are relevant for most projects:
* `flink-table-api-scala`: The Table & SQL API for pure table programs using
the Scala programming language (in early development stage, not recommended!).
* `flink-table-api-java-bridge`: The Table & SQL API with DataStream/DataSet
API support using the Java programming language.
* `flink-table-api-scala-bridge`: The Table & SQL API with DataStream/DataSet
API support using the Scala programming language.
-* `flink-table-planner`: The table program planner and runtime. This was the
only planner of Flink before the 1.9 release. It is still the recommended one.
-* `flink-table-planner-blink`: The new Blink planner.
+* `flink-table-planner`: The table program planner and runtime. This was the
only planner of Flink before the 1.9 release. It's no longer recommended since
Flink 1.11.
+* `flink-table-planner-blink`: The new Blink planner, which has become the
default one since Flink 1.11.
* `flink-table-runtime-blink`: The new Blink runtime.
* `flink-table-uber`: Packages the API modules above plus the old planner into
a distribution for most Table & SQL API use cases. The uber JAR file
`flink-table-*.jar` is located in the `/lib` directory of a Flink release by
default.
* `flink-table-uber-blink`: Packages the API modules above plus the Blink
specific modules into a distribution for most Table & SQL API use cases. The
uber JAR file `flink-table-blink-*.jar` is located in the `/lib` directory of a
Flink release by default.
diff --git a/docs/dev/table/index.zh.md b/docs/dev/table/index.zh.md
index 561890d..e7bb0f5 100644
--- a/docs/dev/table/index.zh.md
+++ b/docs/dev/table/index.zh.md
@@ -38,7 +38,7 @@ Table API 和 SQL 两种 API 是紧密集成的，以及 DataStream 和 DataSet
Planner 的作用主要是把关系型的操作翻译成可执行的、经过优化的 Flink 任务。两种 Planner 所使用的优化规则以及运行时类都不一样。
它们在支持的功能上也有些差异。
-<span class="label label-danger">注意</span> 对于生产环境，我们建议使用在1.9之前就已经存在的 Old
Planner。
+<span class="label label-danger">注意</span> 对于生产环境，我们建议使用在1.11版本之后已经变成默认的Blink
Planner。
所有的 Table API 和 SQL 的代码都在 `flink-table` 或者 `flink-table-blink` Maven artifacts
下。
@@ -49,8 +49,8 @@ Planner 的作用主要是把关系型的操作翻译成可执行的、经过优
* `flink-table-api-scala`: Table 和 SQL API，使用 Scala 语言编写的，给纯 table
程序使用（还在早期开发阶段，不建议使用）
* `flink-table-api-java-bridge`: Table 和 SQL API 结合 DataStream/DataSet API
一起使用，给 Java 语言使用。
* `flink-table-api-scala-bridge`: Table 和 SQL API 结合 DataStream/DataSet API
一起使用，给 Scala 语言使用。
-* `flink-table-planner`: table Planner 和运行时。这是在1.9之前 Flink 的唯一的
Planner，现在仍然建议使用这个。
-* `flink-table-planner-blink`: 新的 Blink Planner。
+* `flink-table-planner`: table Planner 和运行时。这是在1.9之前 Flink 的唯一的
Planner，但是从1.11版本开始我们不推荐继续使用。
+* `flink-table-planner-blink`: 新的 Blink Planner，从1.11版本开始成为默认的 Planner。
* `flink-table-runtime-blink`: 新的 Blink 运行时。
* `flink-table-uber`: 把上述模块以及 Old Planner 打包到一起，可以在大部分 Table & SQL API
场景下使用。打包到一起的 jar 文件 `flink-table-*.jar` 默认会直接放到 Flink 发行版的 `/lib` 目录下。
* `flink-table-uber-blink`: 把上述模块以及 Blink Planner 打包到一起，可以在大部分 Table & SQL API
场景下使用。打包到一起的 jar 文件 `flink-table-blink-*.jar` 默认会放到 Flink 发行版的 `/lib` 目录下。
