service instance disconnected
Describe the bug
The springcloudalibaba service disconnected after successful registration, but the service instance status is normal. And there is no relevant information in the Nacos log
Expected behavior The service instance is started and successfully registered with Nacos, and the health status is always normal
Actually behavior The service instance was started and successfully registered with Nacos. After about 10 seconds, the health status of the service was displayed as an issue on the Nacos client, but there were no abnormalities in the actual service. Additionally, there were services on the same server with normal health status and no disconnected services
- If you register ephemeral instance, you should check your app is running well and no resource problem(such as CPU, FullGC, network problem) so that the heart beat can send to nacos server to keepalive.
- if you register persistent instance, you should check your configured health check policy is right, such as the port, the protocol. What's more, you should make sure your network is well which from nacos server to your app, you can run
telnet ${your_app_ip} ${your_app_health_check_port}command to check the network whether well.
1.The second image is an image using the telnet command 2.In the third image, it can be seen that an instance is always in a normal state, and the excluded instances are all removed at a fixed time after being connected. Is there a version conflict that could cause this phenomenon
I see, It should be your app only register instance, but not send heartbeat to keepalive, so nacos server think your instance has exception and need removed.
You should check your app and find out why your app don't send heartbeat to server.