Data Lake Analytics - Deprecated:FAQ related to the serverless Presto engine

You are not charged for DDL statements executed in the public cluster. If other statements are executed in the public cluster, you are charged based on the pay-per-byte billing type. For more information about the pay-per-byte billing type, see Pay-per-byte.

The public cluster is a cluster that is shared by all users. The computing power provided for each user approximately equals the computing power provided by a cluster with 10 CPU cores.

Cause: If the number of rows returned by the SELECT statement is excessively large, the page of the DLA console may close unexpectedly. The LIMIT 500 clause is added to the SELECT statement to ensure a stable page display of the DLA console.

Solution: To cancel this limit, you can use a MySQL client to connect to DLA.

You must use an Alibaba Cloud account to execute SQL statements in the DLA console. If you use the credentials of a RAM user, you must use the RAM user to create a DLA sub-account in the DLA console. Then, use the username and password of the DLA sub-account to execute SQL statements in the Data Management (DMS) console.

Asynchronous execution is used for SQL statements that are related to extract, transform, load (ETL) tasks, such as INSERTand SELECT statements. The execution of these SQL statements is time-consuming. If you use synchronous execution, the connection between the client and the server is always occupied during this period of time. If a network exception occurs, the query fails. If you use asynchronous execution, an ID is returned immediately after you execute an SQL statement. Then, you can use the SHOW QUERY_TASK statement to view the status of the query.

/*+ run-async=true*/ SELECT * FROM tbl;

Syntax for querying the status of asynchronous execution:

In the left-side navigation pane of the DLA console, choose Serverless Presto > Execute. On the Execute page, enter your SQL statements in the code editor and click Async Execute(F9). Click Result Set to obtain the value of ASYNC_TASK_ID. After that, enter SHOW QUERY_TASK WHERE ID = 'Value of ASYNC_TASK_ID' in the code editor and click Sync Execute(F8). Then, click Result Set to obtain the value of the result_file_oss_file field. This field indicates the OSS directory in which the result data is stored.

• Add a hint to the SQL statement that you want to execute. Sample statement: /*+cluster=dladw*/SELECT * FROM tbl;

• Append @<Cluster ID> to the username in the SQL statement that you want to execute.

• Set the default cluster to dladw.

• Cause: The VPC ID or instance ID that you entered is incorrect.
  Solution: Enter the correct VPC ID and instance ID.

• Cause: Data migration has been performed on your RDS instance. The instance ID that appears in the DLA console is incorrect.

  Solution: Add USE_VPC_CLOUD_INSTANCE_ID = 'true' to the statement that is used to create the database schema. Sample statement:

  CREATE SCHEMA `test_db` WITH DBPROPERTIES (   
  CATALOG = 'mysql',   
  LOCATION = 'jdbc:mysql://rm-xxx.mysql.rds.aliyuncs.com:3306/dla_test',  
  USER='dla_test',   PASSWORD='xxxxx',   INSTANCE_ID = 'rm-xxx',   
  VPC_ID = 'vpc-xxxx',   
  USE_VPC_CLOUD_INSTANCE_ID = 'true' );

You can query the number of files in an OSS table by using the following statement:

In the preceding statement, customer indicates the customer table in the ApsaraDB RDS for MySQL database. DLA uses two data queries that are sent over a JDBC connection to scan the data of the customer table in the ApsaraDB RDS for MySQL database.

-- Original SQL statement   
select * from customer;   
-- SQL statement for subqueries   
select * from customer where id >=0 and id <=10000;   
select * from customer where id >=10000 and id <=20000;

Cause: The data type of underlying data does not map the INT data type that is supported by DLA.

Solution: Use the BIGINT data type to map the data type of underlying data.