Exporting SQL Job Results

Store the data results of analyzed SQL jobs in a specified location in the desired format.

By default, DLI stores SQL job results in its job bucket. You can also download job results to a local host or export job results to a specified OBS bucket.

Exporting Job Results to the DLI Job Bucket

DLI specifies a default OBS bucket for storing job results. You can configure the bucket information on the Global Configuration > Project page of the DLI management console. Once a job is complete, the system automatically stores its results to this bucket.

The following conditions must be met if you want to read job results from the DLI job bucket:

You have configured the job bucket on the Global Configuration > Project page of the DLI management console by referring to Project Configurations.
You have submitted a service ticket to request the whitelisting of the feature that allows writing job results to buckets.
The user who executes jobs has been granted read and write permissions either on the job bucket or on the jobs/result path of the job bucket.
For details, see Creating a Custom Bucket Policy.

For how to obtain job results from the DLI job bucket, see "Object Management" > Downloading Objects in Object Storage Service User Guide.

Exporting Job Results to a Specified Location in Another Bucket

In addition to storing job results in the default bucket, you can also export them to a specified location in another bucket, increasing the flexibility of job result management and making it easier to organize and manage them.

On the console, you can only view a maximum of 1,000 job results. To view additional results, you can export them to an OBS path. The procedure is as follows:

You can export job results on either the SQL Jobs or the SQL Editor page.

SQL Jobs page: In the navigation pane on the left, choose Job Management > SQL Jobs. On the displayed page, locate the row containing a desired job, click More in the Operation column, and select Export Result.
SQL Editor page: In the navigation pane on the left, choose SQL Editor. On the displayed page, once query statements are successfully executed, click next to the View Result tab to export job results.

If there are no numerical columns in the query results, job results cannot be exported.
Ensure that the user who exports job results has the read and write permissions on the OBS bucket.

**Table 1** Parameters
Parameter	Mandatory	Description
Data Format	Yes	Choose a data format for the job results you want to export. The options include json and csv.
Queues	Yes	Select the queue where the job is executed. SQL jobs can only be executed on SQL queues.
Compression Format	No	Compression format of the data to be exported. The options are: none bzip2 deflate gzip
Storage Path	Yes	Path in an OBS bucket where the job results are exported If Export Mode is set to New OBS directory, then You need to manually enter a directory name and ensure that the directory name does not exist. Otherwise, the system returns an error message and the export operation cannot be performed. NOTE: The folder name cannot contain special characters (\/:?"<>\|) and cannot start or end with a period (.). For example, after selecting the storage path obs://bucket/src1/, you need to manually enter a directory name to change the path to obs://bucket/src1/src2/* and ensure that the src2 directory name does not exist under src1. The job result export path is obs://bucket/src1/src2/test.csv. If Export Mode is set to Existing OBS directory (Overwritten), then After selecting a bucket path, the job results are exported to that path. If there are files with the same name, they will be automatically overwritten. For example, if you select obs://bucket/src1/ as the bucket path, then The job result export path is obs://bucket/src1/test.csv.
Export Mode	Yes	New OBS directory If you select this mode, a new folder path is created and the job results are saved to this path. This mode is used when you need to save exported results to a new location, making it easier to manage and track job results. If you select this option, you must manually enter an export directory in Storage Path and ensure that the directory must not exist. If the directory already exists, the system displays an error message and the export operation cannot be performed. Existing OBS directory (Overwritten): When exporting job results, you can choose an existing file path as the output directory. If there is a file with the same name in that path, it will be automatically overwritten by the new exported job result file. This mode is used when you only need to save a single job result file in the same path, and you do not need to keep old job results.
Number of Results	No	Number of results to be exported If you do not specify or set it to 0, all results will be exported.
Table Header	No	Whether the job results to be exported contain table headers

Exporting Job Results to a Local Host

You can download the results of asynchronous DDL and QUERY statements to a local directory. By default, you can download a maximum of 1,000 data records to a local host.

The procedure is as follows:

Locate the row containing a desired job whose asynchronous DDL or QUERY statement has been successfully executed, click More in the Operation column, and select Submit Download Request. In the displayed dialog box, click OK. After a few seconds, the Submit Download Request button would change to Download.
Figure 1 Selecting Submit Download Request
Click Download to download the results to your local host.