tencent cloud

Data Lake Compute

DocumentaçãoData Lake Compute

TDLC Command Line Interface Tool Access

Download
Modo Foco
Tamanho da Fonte
Última atualização: 2026-08-11 11:34:52
TDLC is a client command tool provided by Tencent Cloud Data Lake Compute (DLC). Using the TDLC tool, you can submit SQL and Spark tasks to the DLC data engine.
TDLC is written in Go, based on the Cobra framework, and supports configuring multiple buckets and cross-bucket operations. You can view the usage of TDLC by running ./tdlc [command] --help.

Downloading and Installation

The TDLC Tencent Cloud Command Line Interface provides binary packets for Windows, Mac, and Linux operating systems. You can use it after a simple installation and configuration. Select the download based on your client's operating system type.
Operating System
TDLC Binary Packet Download Link
Windows
Mac
Linux
Rename the downloaded file to tdlc. Open the command line on your client, navigate to the download path, and if you are using Mac/Linux, use the chmod +x tdlc command to grant executable permissions to the file. After ./tdlc is executed, if the following content is displayed successfully, the installation is complete and the tool is ready for use.
Tencentcloud DLC command tools is used to play around with DLC.
With TDLC user can manger engines, execute SQLs and submit Spark Jobs.

Usage:
  tdlc [flags]
  tdlc [command]

Available Commands:
  config      
  help        Help about any command
  spark       Submit spark app to engines. 
  sql         Executing SQL. 
  version     

Flags:
      --endpoint string          Endpoint of Tencentcloud account. (default "dlc.tencentcloudapi.com")
      --engine string            DLC engine. (default "public-engine")
  -h, --help                     help for tdlc
      --region string            Region of Tencentcloud account.
      --resource-group string    DLC resource group name of spark engine.
      --role-arn string          Required by spark jar app.
      --secret-id string         SecretId of Tencentcloud account.
      --secret-key string        SecretKey of Tencentcloud account.
      --token string             Token of Tencentcloud account.

Use "tdlc [command] --help" for more information about a command.

Usage Instructions

Global Parameters

TDLC provides the following global parameters.
Global Parameters
Description
--endpoint string
Service connection address, and dlc.tencentcloudapi.com is used by default.
--engine string
DLC data engine name. The default value is public-engine. It is recommended to use a Dedicated Data Engine.
--resource-group string
(Optional) Resource group name of the DLC Standard Spark engine. The default value is the default resource group default-rg-xxxx.
--region string
Region to be used Examples: ap-nanjing, ap-beijing, ap-guangzhou, ap-shanghai, ap-chengdu, ap-chongqing, na-siliconvalley, ap-singapore, ap-hongkong
--role-arn string
When submitting Spark jobs, you need to specify the permissions to access COS files. This involves assigning the appropriate rolearn with the required permissions. For details on the rolearn, see Configuring Data Access Policy.
--secret-id string
Tencent Cloud account secretId
--secret-key string
Tencent Cloud account secretKey
--token string
(Optional) Tencent Cloud account temporary token

CONFIG Command

The config command can configure commonly used parameters. The configured parameters will be provided as default values. Command line parameters will override the configured config parameters.
Command
Note
list
List current default configurations.
set
Change configuration.
unset
Reset configuration.
Example:
./tdlc config list
./tdlc config set secret-id={1} secret-key={2} region={b}
./tdlc config unset region

SQL Subcommands

SQL subcommands currently support Presto or SparkSQL clusters. The following parameters are supported by SQL subcommands.
Parameter
Note
-e, --exec
Execute SQL statements
-f, --file
Execute SQL files. If there are multiple SQL files, separate them with ;. Local paths and COS paths (cos://{bucket}/{key} or cosn://{bucket}/{key}) are supported.
--no-result
Do not fetch results after execution.
-p, --progress
Display execution progress
-q, --quiet
Quiet mode, do not wait for task execution status after submission.
--conf
Session configuration item in the format of key=value, can be specified multiple times. Any native Spark conf can be passed through when --submit-mode=batch-sql.
--submit-mode
Submission channel. Options: tasks (default) or batch-sql.
--driver-size
Driver specification. Options: small / medium / large / xlarge. For memory-optimized clusters, use m.small / m.medium / m.large / m.xlarge. Takes effect only when --submit-mode=batch-sql.
--executor-size
Executor specification. Same values as --driver-size. Takes effect only when --submit-mode=batch-sql.
--executor-num
Number of executors. Takes effect only when --submit-mode=batch-sql.
--executor-max-num
Upper limit of the number of executors. Dynamic allocation is enabled when it is greater than --executor-num. Takes effect only when --submit-mode=batch-sql.
--param
Business parameter injection in the format of key=value, can be specified multiple times. Used to replace the ${key} placeholder in SQL.
--param-strict
Whether to enable strict mode, defaults to true. When enabled, any unreplaced ${key} placeholder in SQL causes the submission to fail directly before submission.
--task-name
Custom task name, used by downstream systems for display and identification. Applicable to both tasks and batch-sql submission modes. If not specified, no custom task name is set.
Example:
./tdlc sql -e "SELECT 1" --secret-id aa --secret-key bb --region ap-beijing --engine public-engine --resource-group my-rg
./tdlc sql -f ~/biz.sql --no-result 
./tdlc sql -f cosn://bucket/path/biz.sql
./tdlc sql -e "SELECT 1" --submit-mode batch-sql --driver-size small --executor-size small --executor-num 2 --conf spark.sql.adaptive.enabled=true

SPARK Subcommands

Spark subcommands include the following commands, which can be used to submit Spark jobs, view running logs, and terminate tasks.
Command
Description
submit
Submit a task through spark-submit.
run
Execute Spark jobs.
log
View running logs.
list
View the list of Spark jobs.
kill
Terminate tasks
The following parameters are supported by the Spark submit subcommand, with file-related parameters supporting the use of local files or COSN protocol.
Parameter
Note
--driver-size
Driver specification, defaults to small, medium, large, xlarge. For memory-optimized clusters, use m.small, m.medium, m.large, m.xlarge.
--executor-size
Executor specification, defaults to small, medium, large, xlarge. For memory-optimized clusters, use m.small, m.medium, m.large, m.xlarge.
--executor-num
Quantities of executors
--files
Additional dependency files to upload (excluding jar/py files), separated by commas ,. These files can be accessed directly in the working directory of the Driver/Executor at runtime.
--archives
Dependent compressed files
--class
Primary function run by the Java/Scala application.
--jars
Dependent jar files, separated by the comma ,.
--name
Application name
--py-files
Dependent Python files, supports .zip, .egg, .py formats.
--conf
Additional configurations
--network
Network Configuration, such as --network "network name"
--task-name
Custom task name, used by downstream systems for display and identification. If not specified, no custom task name is set. This parameter identifies the task for each run. It does not replace --name, which identifies the Spark application.
Example:
./tdlc spark submit --name spark-demo1 --engine sparkjar --jars /root/sparkjar-dep.jar --class com.demo.Example /root/sparkjar-main.jar arg1
./tdlc spark submit --name spark-demo2 cosn://bucket1/abc.py arg1
./tdlc spark submit --name spark-demo3 --task-name daily-etl-20260730 cosn://bucket1/etl.py
Próximo: Política de privacidade

Ajuda e Suporte

Esta página foi útil?

comentários