After you publish SOFAStack, Dubbo, or Spring Cloud microservices, you can control and administer them in the Service Mesh console.
Service control
Use the Service Catalog page to query information about services, applications, service consumers, and service providers. For more information, see Service Catalog.
Service administration
Service administration includes features such as service throttling, service routing, service circuit breaking, fault injection, and fault isolation.
Service throttling
In high-concurrency scenarios, you can use service throttling to keep requests and concurrent traffic within an acceptable range for your application. This practice helps ensure service stability with existing resources and achieve high availability. For more information, see Service throttling.
Service routing
When a service consumer can connect to multiple service providers, routing rules determine which provider is used. You can define multiple routing rules to handle various scenarios. For more information, see Service routing.
Service circuit breaking
If a server-side interface in a microservice becomes unstable and has frequent timeouts or errors, it can cause a cascade of failures. You can enable service circuit breaking for your application to handle this situation. A circuit breaker makes the faulty service fail fast by immediately returning an error. This practice releases system resources and improves both user experience and system performance. For more information, see Service circuit breaking.
Fault injection
You can use the fault injection feature to introduce faults into a test application. This practice tests how the application handles abnormal conditions. Based on the results, you can adjust your application to improve its robustness in a production environment. For more information, see Fault injection.
Fault isolation
If a faulty service triggers a circuit breaker, the entire service can become unavailable. Fault isolation prevents this by pinpointing the abnormal service instance. It enables fine-grained, instance-level isolation and traffic removal. This contains the fault's impact, making it smaller and more controllable. For more information, see Fault isolation.