tencent cloud

Service Registry and Governance

문서Service Registry and GovernancePolaris ObservabilityMonitoring and Troubleshooting

Monitoring and Troubleshooting

PDF
포커스 모드
폰트 크기
마지막 업데이트 시간: 2026-05-07 18:07:07

Troubleshooting Overview

This article introduces the fundamental troubleshooting approach and operational procedures for using Polaris, to help Ops personnel locate and resolve issues during their use of Polaris.
When your client accesses Polaris and encounters an error, you can view the monitoring of PolarisMesh to preliminarily locate what error has occurred.

1. Find the Interface Reporting Errors

First, based on the client logs, identify which interface of Polaris is being accessed. Then, go to the Polaris monitoring page, select the specific interface, and view the monitoring.

The monitoring and alarm system of PolarisMesh currently supports the following interfaces:
API Name
API Description
POST:/eureka/apps/{application}
eureka client service instance registration.
GET:/eureka/apps
The eureka client queries all service instances.
GET:/eureka/apps/delta
The eureka client incrementally pulls service instances.
GET:/eureka/apps/{application}
The eureka client queries all instances of a service.
GET:/eureka/apps/{application}/{instanceid}
The eureka client queries a specific instance under a service.
PUT:/eureka/apps/{application}/{instanceid}
The eureka client reports service instance heartbeats.
DELETE:/eureka/apps/{application}/{instanceid}
eureka client service instance deregistration.
PUT:/eureka/apps/{application}/{instanceid}/status
eureka client service instance status change.
DELETE:/eureka/apps/{application}/{instanceid}/status
The eureka client deletes service instance changes.
/v1.PolarisGRPC/RegisterInstance
grpc client service instance registration.
/v1.PolarisGRPC/DeregisterInstance
grpc client service instance deregistration.
/v1.PolarisGRPC/Heartbeat
grpc client service instance heartbeat reporting.
/v1.PolarisGRPC/Discover
grpc client service discovery.
For example, use the following command to request Polaris and access the Eureka interface to register an instance.
curl -v -X POST -H "Accept: application/json" -H "Content-type: text/plain"  http://{$polaris_ip}:8761/eureka/apps/test -d '{"instance":{"instanceId":"test-1","securePort":{"$":"0","@enabled":"false"},"ipAddr":"{$client_ip}}","status":"UP","port":{"$":"{$client_port}}","@enabled":"true"},"hostName":"{$client_ip}"}}'
On the monitoring page, you can select the interface POST:/eureka/apps/{application} to view the monitoring information for this request.

2. Locate the Issue Based on Monitoring Information

Based on the Request Failure Breakdown group in the monitoring information, you can preliminarily locate the cause of the error. Currently, it includes the following error types:
Other errors: These are errors where the PolarisMesh service returns an error code of 4XX, indicating a parameter validation failure. You need to check whether the parameters and HTTP headers you passed are correct. If you are using a non-Eureka access method and have logged the response from the PolarisMesh service, you can check the logs. The PolarisMesh response packet contains the reason for the parameter validation failure.
System errors: These are errors where the PolarisMesh service returns an error code of 5XX. The occurrence of such errors indicates an exception in PolarisMesh. You need to submit a ticket and contact Tencent Cloud engineers to resolve the issue.
Network errors: These are errors where the PolarisMesh service returns an error code of negative number. A typical scenario is when the client disconnects while PolarisMesh is writing data to it after the client's data is received, which generates such errors. You need to check your code logic to see whether the connection was closed prematurely.
이전 주제: Business Monitoring Alarm다음 주제: Alarm Configuration

도움말 및 지원

문제 해결에 도움이 되었나요?

피드백