Enterprise Distributed Application Service (EDAS) is a platform as a service (PaaS) platform for cloud-native applications. When your applications do not need full application performance monitoring (APM) or microservices governance -- for example, during development and testing, or when you already use tools like SkyWalking -- paying for bundled ARMS and MSE adds unnecessary cost. Custom combination mode provides EDAS application lifecycle management for free. Activate Application Real-Time Monitoring Service (ARMS) or Microservices Engine (MSE) only when you need them.
How the two modes compare
EDAS offers two activation modes. Use this table to decide which fits your workload.
| Dimension | Integrated mode | Custom combination mode |
|---|---|---|
| What is included | Application lifecycle management + ARMS + MSE | Application lifecycle management only (free) |
| EDAS billing | Subscription or pay-as-you-go | Free (no EDAS charges) |
| ARMS and MSE billing | Included in EDAS fees | Billed separately if activated |
| Application monitoring | Full ARMS APM | Prometheus monitoring only |
| Microservices governance | Full MSE governance | Not included (bring your own) |
| Canary release | Supported | Not supported |
| Auto scaling by response time and QPS | Supported | Not supported (requires ARMS) |
| Registration SLA guarantee | Provided | Not provided unless MSE is activated |
| Best for | Production applications that need APM and governance. We recommend integrated mode for production applications. | Dev/test environments, monolithic apps, or teams with existing monitoring |
The following example shows the pay-as-you-go billing method for the Standard Edition of EDAS:
For example, the pay-as-you-go EDAS Standard Edition costs CNY 2 per application instance per day. This includes the Professional Edition of MSE microservice governance (CNY 0.96/instance/day) and the Basic Edition of ARMS application monitoring (CNY 2.4/instance/day). This saves about 40% compared to purchasing MSE microservice governance and ARMS application monitoring separately.
Capabilities included without ARMS or MSE
The following table lists what EDAS supports in custom combination mode when neither ARMS nor MSE is activated.
| Category | Capability | Details |
|---|---|---|
| Microservices framework | Framework support | Spring Cloud, Dubbo, and High-speed Service Framework (HSF) |
| Application release and operations | Lifecycle management | Create, deploy, roll back, start, stop, restart, delete, and scale up applications |
| Log management | Query online logs and log files; integrates with Log Service | |
| Auto scaling | Manual scaling, scheduled scaling, and metric-based auto scaling. Note Scaling by response time and queries per second (QPS) requires ARMS. See Activate ARMS. | |
| CI/CD | OpenAPI SDK, Jenkins plug-ins integration, and the latest version of Apsara DevOps | |
| Release strategies | At-a-time release and phased release (canary release is not supported) | |
| Registration and configuration center | Distributed registration and configuration centers | |
| Enterprise permission management | Access control | Tenant isolation and Resource Access Management (RAM) policies |
| Multi-environment management | Namespace-based resource isolation for development, test, and production environments | |
| Hybrid cloud | Standard Edition: Not supported. Professional and Platinum Editions: Deploy applications across Elastic Compute Service (ECS) clusters and Kubernetes clusters in hybrid cloud architecture. | |
| Monitoring | Metrics | Prometheus monitoring only |
When to use custom combination mode
Start with custom combination mode in any of these situations:
Development and testing -- Non-production workloads rarely need APM and governance. Custom combination mode eliminates those costs.
Existing monitoring stack -- If your team already uses SkyWalking or another monitoring tool, skip ARMS and connect your own system.
Monolithic applications -- A single application without microservices dependencies does not need MSE governance.
Limits
Custom combination mode is in public preview and available only in the China (Chengdu) region.
Only Kubernetes clusters are supported.
Only Java applications are supported.
Create an application in custom combination mode
Create the application
Log on to the . In the left-side navigation pane, choose Application Management > Applications. In the top navigation bar, select a region and a microservice namespace. Click Create Application.
In the Basic Information step, configure the following parameters, and then click Next. For image-based deployments, see Deploy a Java microservices application with an image in a Kubernetes cluster.
Parameter Description Cluster Type Select Kubernetes Clusters. Hosted Applications Select Java. Select Application Choose one of the following based on your deployment package: - Java: Use a universal JAR package for Dubbo or Spring Boot applications. After selection, set the Java Environment parameter. - Tomcat: Use a universal WAR package for Dubbo or Spring applications. After selection, set the Java Environment and Container Version parameters. - EDAS-Container (HSF): Use a WAR or fat JAR package for HSF applications. After selection, set the Java Environment, Pandora Version, and Ali-Tomcat Version parameters. Monitoring and Governance Solution Select Custom Combination Solution. In the Configurations step, configure the environment information, basic application settings, deployment method, and resource parameters, and then click Next. For details, see Deploy an application with a JAR or WAR package in an ACK cluster.
(Optional) In the Advanced Settings step, configure additional options such as scheduling rules. For details, see Configure scheduling rules.
Click Create Application. In the Creation Completed step, click Create Application again.
In the Confirm Application Change Precheck dialog box, click Start Precheck. After the precheck completes, review the results and click Continue. If you modify any precheck items, click Check Again before proceeding.
Verify the deployment
The deployment takes several minutes to complete. Track progress on the Change List page.
After deployment finishes, go to the Application Overview page and check the pod status. If all pods show a running state, the application is deployed. Click a pod status entry to view Deployments, pod details, and advanced configurations.
What to do next
After deployment, configure load balancing to expose your application:
Internet access -- Add an Internet-facing Server Load Balancer (SLB) instance. See Bind CLB instances.
VPC-internal access -- Add an internal-facing SLB instance so that other nodes in the same Virtual Private Cloud (VPC) can reach the application. See Reuse a CLB instance.