Best Practices

Suggestions on APM Security Configuration

This section provides guidance for enhancing the overall security of APM. You can continuously evaluate the security of APM and combine different security capabilities provided by APM.

Associating Traces with Logs

Associate trace IDs with logs in Log Tank Service (LTS). If a fault occurs, you can quickly find out the logs based on the associated trace IDs for troubleshooting.

Locating Performance Problems Using APM Profiler

Use APM Profiler (a performance profiling tool) to locate the code that consumes excessive resources.

Locating Out of Memory Problems Using Profiler

Use APM Profiler to locate memory overflows.

Locating the Causes of Request Errors

Sudden request increases or load changes may easily cause application performance problems. APM has powerful analysis tools for cloud application diagnosis. It displays application statuses, call processes, and user operations through topologies and tracing, so that you can quickly locate and resolve faults and performance bottlenecks.

Embedding APM Pages into a Self-built System

APM pages can be embedded into a self-built system. Specifically, create a custom identity broker through the federation proxy mechanism of Identity and Access Management (IAM) and embed a login link into your self-built system. You can then check APM pages on your system without the need to log in to Huawei Cloud.

Encrypting AK/SK for Deploying an APM Agent in a CCE Container

Encrypt the AK/SK when deploying an APM Agent on CCE.

Searching for Span Information

In the distributed architecture, the calls between microservices are complex. If it takes much time to respond to external requests or some requests become abnormal, you can specify a trace ID or set other criteria on the Tracing page to check trace details.

Connecting to APM Using OpenTelemetry

Huawei Cloud APM is compatible with OpenTelemetry and can directly receive the trace data reported using the OpenTelemetry SDK or Agent.