The photo/video publishing service started processing requests slowly during a sharp traffic increase. How to determine whether the problem is on your side or with external integration?
sobes.tech AI
Answer from AI
To determine exactly where the performance issue occurs — on your side or on the external integration side — you need to conduct a system analysis and monitoring:
-
Logging and metrics. Check your service response time and the response wait time from external integrations. If the delay occurs during external API calls, the problem is likely there.
-
Request tracing. Use distributed tracing (e.g., Jaeger or Zipkin) to trace the request path and identify bottlenecks.
-
Load testing. Conduct tests on your side without calling external services to see how your service handles load.
-
Resource monitoring. Check CPU, memory, network, and disk usage on your servers — the bottleneck might be within the infrastructure.
-
Queue and timeout analysis. If external integration responds with delay or errors, this will be visible in logs and metrics.
Thus, by combining monitoring, logging, and testing, you can localize the problem and understand where exactly to optimize the work.