Data Extraction

Last updated: 2020-10-27 17:14:55


    COS Select allows you to filter out desired data at the storage level, significantly reducing the amount of data transferred by COS, thereby lowering your usage costs, and improving data acquisition efficiency. In the COS console, you can extract data from individual files stored in buckets using the standard SQL templates we provide or by specifying statements that comply with syntax rules.


    • COS Select currently supports objects in JSON or CSV format only in public cloud regions (Chinese mainland), and objects in Parquet format only in Beijing region.
    • Please make sure the file to be extracted complies with COS Select specifications. For more information on COS Select specifications, see SELECT Overview.
    • COS Select in the console supports data extraction of up to 40 M from 128 MB files. To process larger files or extract more data, please use the API or SDKs.


    1. Log in to the COS Console.
    2. On the left sidebar, click Bucket List.
    3. Click the bucket name you want to enter the bucket.
    4. Under File List, locate the object to extract data from, and click More Actions > Extract under the "Actions" column.


      Currently, COS Select can extract data only from objects in STANDARD and STANDARD_IA storage classes.

    5. Configure the fields File Type, Header Field, Separator, Compression Format, and Export Format.
    6. Click Select an SQL Template, select the desired template statement, and click OK.
    7. Edit the statement in the text box based on your actual file and then click Run SQL.
    8. Now you can see the first 100 results in the text box at the bottom. To obtain the complete set of data, click Export.

    Was this page helpful?

    Was this page helpful?

    • Not at all
    • Not very helpful
    • Somewhat helpful
    • Very helpful
    • Extremely helpful
    Send Feedback