Submitting a SQL Job (Recommended)

Function

This API is used to submit jobs to a queue using SQL statements.

The job types support DDL, DCL, IMPORT, QUERY, and INSERT. The IMPORT function is the same as that described in Importing Data (Deprecated). The difference lies in the implementation method.

Additionally, you can use other APIs to query and manage jobs. For details, see the following sections:

• Querying Job Status

• Querying Job Details

• Exporting Query Results

• Querying All Jobs

• Canceling a Job (Recommended)

Note

This API is synchronous if job_type in the response message is DCL.

URI

• URI format

  POST /v1.0/{project_id}/jobs/submit-job

• Parameter description

  Table 1 URI parameter

  Parameter	Mandatory	Type	Description
  project_id	Yes	String	Project ID, which is used for resource isolation. For details about how to obtain its value, see Obtaining a Project ID.

Request Parameters

Table 2 Request parameters

Parameter	Mandatory	Type	Description
sql	Yes	String	SQL statement that you want to execute.
currentdb	No	String	Database where the SQL statement is executed. This parameter does not need to be configured during database creation.
queue_name	No	String	Name of the queue to which a job to be submitted belongs. The name can contain only digits, letters, and underscores (_), but cannot contain only digits or start with an underscore (_).
conf	No	Array of strings	You can set the configuration parameters for the SQL job in the form of Key/Value. For details about the supported configuration items, see Table 3.
tags	No	Array of objects	Label of a job. For details, see Table 4.
engine_type	No	String	Type of the engine that executes jobs.

Table 3 Configuration parameters description

Parameter	Default Value	Description
spark.sql.files.maxRecordsPerFile	0	Maximum number of records to be written into a single file. If the value is zero or negative, there is no limit.
spark.sql.autoBroadcastJoinThreshold	209715200	Maximum size of the table that displays all working nodes when a connection is executed. You can set this parameter to -1 to disable the display. Note Currently, only the configuration unit metastore table that runs the ANALYZE TABLE COMPUTE statistics noscan command and the file-based data source table that directly calculates statistics based on data files are supported.
spark.sql.shuffle.partitions	200	Default number of partitions used to filter data for join or aggregation.
spark.sql.dynamicPartitionOverwrite.enabled	false	Whether DLI overwrites the partitions where data will be written into during runtime. If you set this parameter to false, all partitions that meet the specified condition will be deleted before data overwrite starts. For example, if you set false and use INSERT OVERWRITE to write partition 2021-02 to a partitioned table that has the 2021-01 partition, this partition will be deleted. If you set this parameter to true, DLI does not delete partitions before overwrite starts.
spark.sql.files.maxPartitionBytes	134217728	Maximum number of bytes to be packed into a single partition when a file is read.
spark.sql.badRecordsPath	-	Path of bad records.
dli.sql.sqlasync.enabled	true	Indicates whether DDL and DCL statements are executed asynchronously. The value true indicates that asynchronous execution is enabled.
dli.sql.job.timeout	-	Sets the job running timeout interval. If the timeout interval expires, the job is canceled. Unit: second
spark.sql.legacy.correlated.scalar.query.enabled	false	If set to true: When there is no duplicate data in a subquery, executing a correlated subquery does not require deduplication from the subquery's result. If there is duplicate data in a subquery, executing a correlated subquery will result in an error. To resolve this, the subquery's result must be deduplicated using functions such as max() or min(). If set to false: Regardless of whether there is duplicate data in a subquery, executing a correlated subquery requires deduplicating the subquery's result using functions such as max() or min(). Otherwise, an error will occur.
spark.sql.optimizer.dynamicPartitionPruning.enabled	true	This parameter is used to control whether to enable dynamic partition pruning. Dynamic partition pruning can help reduce the amount of data that needs to be scanned and improve query performance when executing SQL queries. When set to true, dynamic partition pruning is enabled. SQL automatically detects and deletes partitions that do not meet the WHERE clause conditions during query. This is useful for tables that have a large number of partitions. If SQL queries contain a large number of nested left join operations and the table has a large number of dynamic partitions, a large number of memory resources may be consumed during data parsing. As a result, the memory of the driver node is insufficient and there are frequent Full GCs. To avoid such issues, you can disable dynamic partition pruning by setting this parameter to false. However, disabling this optimization may reduce query performance. Once disabled, Spark does not automatically prune the partitions that do not meet the requirements.

Table 4 tags parameters

Parameter	Mandatory	Type	Description
key	Yes	String	Tag key Note A tag key can contain a maximum of 128 characters. Only letters, numbers, spaces, and special characters (_.:+-@) are allowed, but the value cannot start or end with a space or start with _sys_.
value	Yes	String	Note A tag value can contain a maximum of 255 characters. Only letters, numbers, spaces, and special characters (_.:+-@) are allowed.

Response Parameters

Table 5 Response parameters

Parameter	Mandatory	Type	Description
is_success	Yes	Boolean	Indicates whether the request is successfully sent. Value true indicates that the request is successfully sent.
message	Yes	String	System prompt. If execution succeeds, the parameter setting may be left blank.
job_id	Yes	String	ID of a job returned after a job is generated and submitted by using SQL statements. The job ID can be used to query the job status and results.
job_type	Yes	String	Job type. The options include: DDL DCL IMPORT EXPORT QUERY INSERT
schema	No	Array of Map	If the statement type is DDL, the column name and type of DDL are displayed.
rows	No	Array of objects	When the statement type is DDL and dli.sql.sqlasync.enabled is set to false, the execution results are returned directly. However, only a maximum of 1,000 rows can be returned. If there are more than 1,000 rows, obtain the results asynchronously. That is, when submitting the job, set xxxx to true, and then obtain the results from the job bucket configured by DLI. The path of the results on the job bucket can be obtained from the result_path in the return value of the ShowSqlJobStatus API. The full data of the results will be automatically exported to the job bucket.
job_mode	No	String	Job execution mode. The options are as follows: async: asynchronous sync: synchronous

Example Request

Submit a SQL job. The job execution database and queue are db1 and default, respectively. Then, add the tags workspace=space1 and jobName=name1 for the job.

{
    "currentdb": "db1",
    "sql": "desc table1",
    "queue_name": "default",
    "conf": [
        "dli.sql.shuffle.partitions = 200"
    ],
    "tags": [
            {
              "key": "workspace",
              "value": "space1"
             },
            {
              "key": "jobName",
              "value": "name1"
             }
      ]
}

Example Response

{
  "is_success": true,
  "message": "",
  "job_id": "8ecb0777-9c70-4529-9935-29ea0946039c",
  "job_type": "DDL",
  "job_mode":"sync",
  "schema": [
    {
      "col_name": "string"
    },
    {
      "data_type": "string"
    },
    {
      "comment": "string"
    }
  ],
  "rows": [
    [
      "c1",
      "int",
      null
    ],
    [
      "c2",
      "string",
      null
    ]
  ]
}

Status Codes

Table 6 describes the status codes.

Table 6 Status codes

Status Code	Description
200	Submitted successfully.
400	Request error.
500	Internal service error.

Error Codes

If an error occurs when this API is invoked, the system does not return the result similar to the preceding example, but returns the error code and error information. For details, see Error Codes.

last updated: 2025-06-16 14:07 UTC - commit: 2d6c283406071bb470705521bc41e86fa3400203