[ 
https://issues.apache.org/jira/browse/FLINK-21091?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

godfrey he updated FLINK-21091:
-------------------------------
    Description: 
Since FLINK-20435 is finished, an ExecNode graph can be serialized to a json 
plan or a json plan can deserialized to an ExecNode graph. Only Blink planner 
introduce ExecNode, so legacy planner does not supports this feature.

We will introduce the following methods to support json plan operations:
{code:java}
// Get the json plan for the given statement.
String getJsonPlan(String stmt);

//  Returns the execution plan for the given json plan.
String explainJsonPlan(String jsonPlan, ExplainDetail... extraDetails);

//  Execute the given json plan, and return the execution result. 
TableResult executeJsonPlan(String jsonPlan);
{code}

Currently, these methods are *experimental* features, so we introduce them into 
TableEnvironmentInternal as first step.

This is an umbrella issue, we will create more related sub-tasks, including:
1. introduce the above methods
2. support json serialization/deserialization for all StreamExecNodes
3. introduce complete testing to verify ExecNode serialization/deserialization 
for different queries


  was:
Since FLINK-20435 is finished, an ExecNode graph can be serialized to a json 
plan or a json plan can deserialized to an ExecNode graph. Only Blink planner 
introduce ExecNode, so legacy planner does not supports this feature.

We will introduce the following methods to support json plan operation:
{code:java}
// Get the json plan for the given statement.
String getJsonPlan(String stmt);

//  Returns the execution plan for the given json plan.
String explainJsonPlan(String jsonPlan, ExplainDetail... extraDetails);

//  Execute the given json plan, and return the execution result. 
TableResult executeJsonPlan(String jsonPlan);
{code}

Currently, these methods are *experimental* features, so we introduce them into 
TableEnvironmentInternal as first step.

This is an umbrella issue, we will create more related sub-tasks, including:
1. introduce the above methods
2. support json serialization/deserialization for all StreamExecNodes
3. introduce complete testing to verify ExecNode serialization/deserialization 
for different queries



> Support ExecNode json serialization/deserialization
> ---------------------------------------------------
>
>                 Key: FLINK-21091
>                 URL: https://issues.apache.org/jira/browse/FLINK-21091
>             Project: Flink
>          Issue Type: New Feature
>          Components: Table SQL / Planner
>            Reporter: godfrey he
>            Priority: Major
>
> Since FLINK-20435 is finished, an ExecNode graph can be serialized to a json 
> plan or a json plan can deserialized to an ExecNode graph. Only Blink planner 
> introduce ExecNode, so legacy planner does not supports this feature.
> We will introduce the following methods to support json plan operations:
> {code:java}
> // Get the json plan for the given statement.
> String getJsonPlan(String stmt);
> //  Returns the execution plan for the given json plan.
> String explainJsonPlan(String jsonPlan, ExplainDetail... extraDetails);
> //  Execute the given json plan, and return the execution result. 
> TableResult executeJsonPlan(String jsonPlan);
> {code}
> Currently, these methods are *experimental* features, so we introduce them 
> into TableEnvironmentInternal as first step.
> This is an umbrella issue, we will create more related sub-tasks, including:
> 1. introduce the above methods
> 2. support json serialization/deserialization for all StreamExecNodes
> 3. introduce complete testing to verify ExecNode 
> serialization/deserialization for different queries



--
This message was sent by Atlassian Jira
(v8.3.4#803005)

Reply via email to