Docs

Volcano Engine Function as a Service

Microservices applications

Creating microservices apps

Microservices apps can host containerized applications built with any programming language and framework. You can use two deployment options for microservices apps: code package-based deployment and container image-based deployment. Microservices apps support auto scaling and are billed on a pay-as-you-go basis. They apply to scenarios including microservice workloads, MQ consumer services, and other use cases. This topic describes how to create a microservices app.

Prerequisites

veFaaS has been activated.

Cross-service authorization has been completed if you use veFaaS for the first time.

Code development has been completed locally. For details, refer to Native runtime.

(Optional) The Volcano Engine VPC service has been activated, and a VPC, subnets, and security groups have been created in the region where the function is to be deployed. For details, see Creating VPCs, Creating subnets, and Creating security groups.

(Optional) To upload code via TOS, store the code ZIP package in TOS in advance. For specific steps, see Getting started with TOS.

(Optional) If you need to upload code from your on-premises computer or via TOS, make sure that the code ZIP file is not encrypted and its size does not exceed 256 MiB.

(Optional) The Volcano Engine Torch Log Service (TLS) has been activated. If you need to use a custom logging project to handle function logs, ensure that the project and topics have already been created. For details, see Creating log projects and Creating log topics.

(Optional) If you need to manage function resources by using projects, ensure that you have created a project in advance. For details, see Creating projects.

Limits

Microservices apps do not support online function tests or triggers.

For container image-based deployment:

The container image must be pushed to a Volcano Engine CR instance that is deployed in the same region as the function.

After the function is created, wait for the image synchronization to complete before releasing the function.

Procedure

In the top navigation pane, select the target region.

In the left navigation pane, click Function > Function management. On the Function page, click Create function.

On the Select function creation method page, click Microservice function > Create Microservice function. Configure the parameters based on the following descriptions.

Basic information

Parameter	Description
Name	Specify a custom name for the microservices app. The name must be unique within the same account. It cannot be modified after the function is created.
Description	Enter a description. It can be left blank.

Function code

Parameter	Description
Runtime	Select a runtime environment for the function code. In this scenario, only Native runtime is supported.
Deployment method	You can select one of the following deployment methods based on your business requirements. Image: Deploy the function by using a container image. tip Public images are contributed by community developers. veFaaS does not maintain them. If you encounter issues during software usage, provide feedback to the community. CR: Select a specific image for deployment from Volcano Engine CR. You can filter out a specific image by instance, namespace, and repository type (public or private). When you select an image from Container Registry, veFaaS automatically obtains the configurations of the image, including the startup command and environment variables. You can specify whether to apply the configurations of the selected image to the function. Public image: An out-of-the-box preset image of veFaaS. Code tip The code ZIP file must not be encrypted and must not exceed 256 MiB in size. Function template: The built-in, ready-to-use code templates provided by veFaaS. Click Code preview to view the resource files and code details included in the current template. Select an appropriate code template based on your business requirements. Upload code locally: Upload a local ZIP file. Upload from TOS: In the upload area, select the target bucket and ZIP file.
Webserver mode	No. You do not need to include an HTTP Server or expose a port (set Port to -1). After a request is successfully processed, the process exits proactively or automatically upon timeout.
Startup command	Enter the startup command of the application. When the deployment method is set to Image, if you need to specify a script file, use its absolute path and make sure that the script has executable permissions. If you do not specify a script, the /opt/application/run.sh script is executed by default. When the deployment method is set to Code, if you need to specify a script file, use its path relative to the code package root directory and ensure that the script has executable permissions. If you do not enter a script, the script ./run.sh in the root directory of the code package will be executed by default.

Advanced settings

For details, see Configuring functions.

Click Create to create the function.

What to do next

Release the code and configuration of the function. For details, see Releasing functions.

Last updated: 2026.08.19 14:06:05