Updated on 2022-12-07 GMT+08:00

Importing Data

Function

This API is used to import data from a file to a DLI or OBS table. Currently, only OBS data can be imported to a DLI or OBS table.

  • This API is asynchronous.
  • When importing data, you can select an existing OBS bucket path or create an OBS bucket path, but only one OBS bucket path can be specified.
  • If you need to create an OBS bucket, ensure that the bucket name complies with the following naming rules:
    • The name must be globally unique in OBS.
    • The name must contain 3 to 63 characters. Only lowercase letters, digits, hyphens (-), and periods (.) are allowed.
    • The name cannot start or end with a period (.) or hyphen (-), and cannot contain two consecutive periods (.) or contain a period (.) and a hyphen (-) adjacent to each other.
    • The name cannot be an IP address.
    • If the name contains any period (.), the security certificate verification may be triggered when you access the bucket or objects in the bucket.
  • If the type of a column in the source file to be imported does not match that of the target table, the query result of the row will be null.
  • Two or more concurrent tasks of importing data to the same table are not allowed.

URI

  • URI format

    POST /v1.0/{project_id}/jobs/import-table

  • 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

Table 2 Request parameters

Parameter

Mandatory

Type

Description

data_path

Yes

String

Path to the data to be imported. Currently, only OBS data can be imported.

data_type

Yes

String

Type of the data to be imported. Currently, data types of CSV, Parquet, ORC, JSON, and Avro are supported.

NOTE:

Data in Avro format generated by Hive tables cannot be imported.

database_name

Yes

String

Name of the database where the table to which data is imported resides.

table_name

Yes

String

Name of the table to which data is imported.

with_column_header

No

Boolean

Whether the first line of the imported data contains column names, that is, headers. The default value is false, indicating that column names are not contained. This parameter can be specified when CSV data is imported.

delimiter

No

String

User-defined data delimiter. The default value is a comma (,). This parameter can be specified when CSV data is imported.

quote_char

No

String

User-defined quotation character. The default value is double quotation marks ("). This parameter can be specified when CSV data is imported.

escape_char

No

String

User-defined escape character. The default value is a backslash (\). This parameter can be specified when CSV data is imported.

date_format

No

String

Specified date format. The default value is yyyy-MM-dd. For details about the characters involved in the date format, see Table 3. This parameter can be specified when data in the CSV or JSON format is imported.

bad_records_path

No

String

Bad records storage directory during job execution. After configuring this item, the bad records is not imported into the target table.

timestamp_format

No

String

Specified time format. The default value is yyyy-MM-dd HH:mm:ss. For definitions about characters in the time format, see Table 3. This parameter can be specified when data in the CSV or JSON format is imported.

queue_name

No

String

Name of the queue that is specified to execute a task. If no queue is specified, the default queue is used.

overwrite

No

Boolean

Whether to overwrite data. The default value is false, indicating appending write. If the value is true, it indicates overwriting.

partition_spec

No

Object

Partition to which data is to be imported.

  • If this parameter is not set, the entire table data is dynamically imported. The imported data must contain the data in the partition column.
  • If this parameter is set and all partition information is configured during data import, data is imported to the specified partition. The imported data cannot contain data in the partition column.
  • If not all partition information is configured during data import, the imported data must contain all non-specified partition data. Otherwise, abnormal values such as null exist in the partition field column of non-specified data after data import.

conf

No

Array of Strings

User-defined parameter that applies to the job. Currently, dli.sql.dynamicPartitionOverwrite.enabled can be set to false by default. If it is set to true, data in a specified partition is overwritten. If it is set to false, data in the entire DataSource table is dynamically overwritten.

NOTE:

For dynamic overwrite of Hive partition tables, only the involved partition data can be overwritten. The entire table data cannot be overwritten.

Table 3 Definition of characters involved in the date and time patterns

Character

Date or Time Element

Example

G

Epoch ID

AD

y

Year

1996; 96

M

Month

July; Jul; 07

w

Which week in a year

27 (Week 27 in the year)

W

Which week in a month

2 (Second week in the month)

D

Which day in a year

189 (Day 189 in the year)

d

Which day in a month

10 (Day 10 in the month)

u

Which day in a week

1 (Monday), ..., 7 (Sunday)

a

am/pm flag

pm (Afternoon)

H

Hour time (0-23)

2

h

Hour time (1-12)

12

m

Minute time

30

s

Second time

55

S

Which milliseconds

978

z

Time zone

Pacific Standard Time; PST; GMT-08:00

Response

Table 4 Response parameters

Parameter

Mandatory

Type

Description

is_success

No

Boolean

Indicates whether the request is successfully sent. Value true indicates that the request is successfully sent.

message

No

String

System prompt. If execution succeeds, the parameter setting may be left blank.

job_id

No

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_mode

No

String

Job execution mode. The options are as follows:

  • async: asynchronous
  • sync: synchronous

Example Request

{
    "data_path": "obs://home/data1/DLI/t1.csv",
    "data_type": "csv",
    "database_name": "db2",
    "table_name": "t2",
    "with_column_header": false,
    "delimiter": ",",
    "quote_char": ",",
    "escape_char": ",",
    "date_format": "yyyy-MM-dd",
    "timestamp_format": "yyyy-MM-dd'T'HH:mm:ss.SSSZZ",
    "queue_name": "queue2",
    "overwrite": false,
    "partition_spec":{
      "column1":  "2020-01-01",
      "column2":  "columnPartValue"  
     }
}

Example Response

{
  "is_success": true,
  "message": "import data to table t2 started",
  "job_id": "6b29eb77-4c16-4e74-838a-2cf7959e9202",
  "job_mode":"async"
}

Status Codes

Table 5 describes the status code.

Table 5 Status codes

Status Code

Description

200

Import succeeded.

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 Code.