Status
Current state: ["Under Discussion"]
Discussion thread: here (<- link to https://mail-archives.apache.org/mod_mbox/flink-dev/)
JIRA:
Released: 1.16
Please keep the discussion on the mailing list rather than commenting on the wiki (wiki discussions get unwieldy fast).
Motivation
The current syntax/features of Flink SQL is very perfect in both stream mode and batch mode.
But there are still some usability to improve.
for example, If the user wants to insert data into a new table, 2 steps are required:
First, prepare the DDL statement of the table named t1;
Second, insert the data into t1;
These two steps seem to be normal, but if there are many fields, spelling DDL statements can be difficult,
and write out these columns in the following insert statement.
Therefore, we can support CTAS (CREATE TABLE AS SELECT) like MySQL, Oracle, Microsoft SQL Server, Hive, Spark, etc ...
It will be more user friendly. In addition, the Hive dialect already has some support for CTAS.
Proposed Changes
Create table must go through catalog, and in memory catalog is not support CTAS, must be a external catalog.
We can think that there are two types of catalogs in Flink, in-memory catalogs and external catalogs:
In-memory catalog:
- Metadata is a copy of the metadata of the external system, and the user ensures that the entity exists in the external system and the metadata is consistency, otherwise, throw exception when running. CTAS need create table first, so it is hard to ensures that the entity exists in the external system and the metadata is consistency.
- Requires the user to configure some information about the external system through the with syntax, Flink can't get it through in-memory catalog.
External catalog:
I suggest introducing a CTAS clause with a following syntax:
CREATE TABLE [ IF NOT EXISTS ] table_name [ WITH ( table_properties ) ] [ AS query_expression ]
Example:
CREATE TABLE ctas_hudi WITH ('connector.type' = 'hudi') AS SELECT id, name, age FROM hive_catalog.default.test WHERE mod(id, 10) = 0;
Resulting table equivalent to:
CREATE TABLE ctas_hudi ( id BIGINT, name STRING, age INT ) WITH ('connector.type' = 'hudi'); INSERT INTO ctas_hudi SELECT id, name, age FROM hive_catalog.default.test WHERE mod(id, 10) = 0;
Program research
I investigated other bigdata engine implementations such as hive, spark:
Hive(MR) :atomic
Hive MR is client mode, the client is responsible for parsing, compiling, optimizing, executing, and finally cleaning up.
Hive executes the CTAS command as follows:
- Execute query first, and write the query result to the temporary directory.
- If all MR tasks are executed successfully, then create a table and load the data.
- If the execution fails, the table will not be created.
Spark(DataSource v1) : non-atomic
There is a role called driver in Spark, the driver is responsible for compiling tasks, applying for resources, scheduling task execution, tracking task operation, etc.
Spark executes CTAS steps as follows:
- Create a sink table based on the schema of the query result.
- Execute the spark task and write the result to a temporary directory.
- If all Spark tasks are executed successfully, use the Hive API to load data into the sink table created in the first step.
- If the execution fails, driver will drop the sink table created in the first step.
Spark(DataSource v2, Not yet completed, Hive Catalog is not supported yet) : optional atomic
Non-atomic
Non-atomic implementation is consistent with DataSource v1 logic. For details, see CreateTableAsSelectExec .
Atomic
Atomic implementation( for details, see AtomicCreateTableAsSelectExec), supported by StagingTableCatalog and StagedTable .
StagedTable supports commit and abort.
StagingTableCatalog is in memory, when executes CTAS steps as follows:
- Create a StagedTable based on the schema of the query result, but it is not visible in the catalog.
- Execute the spark task and write the result into StagedTable.
- If all Spark tasks are executed successfully, call StagedTable#commitStagedChanges(), then it is visible in the catalog.
- If the execution fails, call StagedTable#abortStagedChanges().
Implementation Plan
Supported Job Mode
Support both streaming and batch mode.
In order to guarantee atomicity, there will be differences in implementation details.
Streaming
Since streaming job are long-running, the table needs to be created first.
- Create the sink table in the catalog based on the schema of the query result.
- Start the job and write the result to the sink table.
Batch
The batch job will end. In order to guarantee atomicity, we usually write the results in a temporary directory.
We will refer to spark DataSource v1 implementation.
Steps:
- Create the sink table in the catalog based on the schema of the query result.
- Start the job and write the result to a temporary directory.
- If the job executes successfully, then load data into the sink table.
- If the job execution fails, then drop the sink table.(This capability requires runtime module support, such as hook, and SQL passes relevant parameters to the runtime module.)
Drop the table if the job fails requires some additional support:
- TableSink needs to provide the CleanUp API, developers implement as needed. Do nothing by default. If an exception occurs, can use this API to drop table or delete the temporary directory, etc.
Precautions
when need drop table:
- User manually cancel the job.
- Job final FAILED status, such as after exceeds the maximum number of task Failovers.
Drop table and TableSink are strongly bound:
Do not do drop table operations in the framework, drop table is implemented in TableSink according to the needs of specific TableSink.
Support in Table API
The executeSql method will be reused
Compatibility, Deprecation, and Migration Plan
It is a new feature with no implication for backwards compatibility.
Test Plan
changes will be verified by UT
Rejected Alternatives
N/A