tencent cloud

SCF CLI allows you to easily deploy a local SCF project to the cloud and view its execution logs on the local CLI by running simple commands.

Function Deployment

SCF CLI uses the deploy sub-command to package a function for deployment. Based on the specified function template configuration file, SCF CLI deploys the specified code package and information in the file, such as function configuration, to the cloud or update a function already in the cloud.

The execution of scf deploy command is fully based on the function template configuration file. For more information on the template file and how to write it, see Template file description.

Parameter Description

The parameters supported by the scf deploy command are as follows:

Usage Examples

Go to the function project directory and run the following command to deploy the package.

$ cd /Users/xxx/code/scf/testproject

$ scf deploy
Compress function './.tcf_build/default-test-func-latest.zip' to zipfile 'default-test-func-latest.zip' success
Deploy namespace 'default' begin
Deploy function 'testproject' success
Deploy namespace 'default' end

After successful deployment, you can query the created "testproject" function in the function list in the console.

If you need to modify the code, you should deploy it again after modification. If the template configuration file is not modified, the function will be updated.

The example below shows how to package and re-upload the code after modification:

$ cd testproject
$ vim index.py

$ scf deploy
Deploy  function 'testproject' success

Ignoring upload

In real projects, you can customize the file content that you don't want to upload, which will be ignored by SCF CLI during packaging.

1. You need to create an ignore folder under the code path.
2. Go to the ignore folder, create a FUNCTIONNAME.ignore configuration file, and describe the content to be ignored in the file.

   Path rule: Define the location of the content to be ignored based on the CodeUri path in template.yaml.

   As shown below, the function name defined in template.yaml is hello, and CodeUri is ./.

   Resources
   default:
    Type: TencentCloud::Serverless::Namespace
    hello:
      Type: TencentCloud::Serverless::Function
      ```properties
        CodeUri: ./
        Type: Event
        Description: This is a template function
        Handler: index.main_handler
        "MemorySize": 128,
        "Runtime": "Python2.7"
        "Timeout": 3,

   The directory hierarchy and HELLO.ignore are as shown below:
   
   After the configuration is completed, the testmodule directory and all .md files under the current path will be ignored during the final upload.

Code upload methods

Currently, SCF CLI supports the following two upload methods:

• Upload code via COS
• Upload code via a local zip package

Uploading code via COS

Using COS can speed up function deployment by up to 80%, significantly boosting your productivity. However, COS may incur charges if the deployment frequency is high and the deployed packages are large. A promotion campaign for SCF and COS has been launched for a limited time. You can claim vouchers when enabling deployment via COS. For more information on the campaign, go to the SCF Console.

Automatic Deployment

You can run the scf configure set --using-cos y command to enable automatic upload via COS. During deployment, the system will automatically enable COS and store the deployment packages under your account.

Manually Specifying

When using the deploy command, you can upload a zip package to the specified bucket by specifying the cos bucket parameter. During deployment, the cloud system will pull the code package from the COS bucket.

Usage Examples

This example shows how to create a testproject project in the /Users/xxx/code/scf directory in Python 2.7.

1. Run the following command to enter the corresponding directory, package the configuration file, and specify the COS bucket as "temp-code-1253970226".

   The COS bucket must be in the same region as the function.

   $ cd /Users/xxx/code/scf/testproject
   $ scf deploy --cos-bucket temp-code-1253970226
   Compress function './.tcf_build/default-testproject-latest.zip' to zipfile 'default-testproject-latest.zip' success
   2019-07-17 10:42:20 put object, url=:https://temp-code-1253970226.cos.ap-shanghai.myqcloud.com/default-testproject-latest-2019-07-17-10-42-20.zip ,headers=:{'x-cos-acl': 'public-read', 'Content-Type': 'application/x-zip-compressed'}
   Upload function zip file 'default-testproject-latest-2019-07-17-10-42-20.zip' to COS bucket 'temp-code-1253970226' success

   After being notified of the locally created zip file based on the returned message, SCF will pass in the zip package to the COS bucket. The cloud system will then pull the code package from the COS bucket and deploy it successfully.

Uploading code via local zip package

If you create a function by uploading the code via local zip package, the zip package size cannot exceed 30 MB; otherwise, the function cannot be created directly. If the package is greater than 30 MB, we recommend you create the function by uploading the code via COS.

The samples in the deploy sub-command are all completed via local zip package uploads. When you use this method, you don't need to include the cos-bucket parameter in the deploy sub-command. SCF CLI will package the local code, generate a zip file with a random name, and then deploy it to the cloud.