When your application experiences latency spikes or error surges, you need to determine which caller is responsible. The Upstream application tab in Enterprise Distributed Application Service (EDAS) tracks response time, request count, and error count for every service that sends data to your application. Use these metrics to isolate whether a performance issue originates from a specific caller, a traffic surge, or a shared downstream dependency.
Open the Upstream application tab
Log on to the EDAS console.
In the left-side navigation pane, choose Application Management > Applications.
On the Applications page, configure the following filters in the top navigation bar:
Select a region.
From the Microservices Namespace drop-down list, select a microservices namespace.
From the Cluster Type drop-down list, select ECS Clusters.
Click the name of the application that you want to manage.
In the left-side navigation pane, choose Application Monitoring > Application Details.
On the Application Details page, select a node on the left side, then click the Upstream application tab on the right side.
Available metrics
The Upstream application tab displays three time-series charts. Each chart plots data for all detected upstream applications over the selected time period.
| Chart | Measures | Diagnostic use |
|---|---|---|
| Response Time | Duration of requests from each upstream application | Detect latency spikes and compare performance across callers |
| Requests | Number of requests sent by each upstream application | Identify traffic patterns and spot unexpected callers |
| Errors | Number of errors from each upstream application | Triage failures by caller and correlate errors with deployments |
Response time
The Response Time chart shows how long each upstream application's requests take to complete.
Detect latency spikes. A sudden increase in response time from a specific upstream application may indicate that the caller is sending more complex requests, or that your application struggles to process requests from that caller.
Compare across callers. If one upstream application consistently shows higher response time than others, investigate whether its request patterns differ or whether a shared downstream dependency is the bottleneck.
Requests
The Requests chart shows the number of requests sent by each upstream application.
Identify traffic patterns. Correlate traffic spikes with response time increases to determine whether latency issues are load-related.
Spot unexpected callers. A new upstream application appearing in the chart may indicate an unplanned integration or a misconfigured service.
Errors
The Errors chart shows the number of errors from each upstream application.
Triage failures by caller. If errors spike for a single upstream application, the issue likely originates from that caller's request format or payload rather than from your application.
Correlate errors with deployments. Compare error spikes against recent deployment timestamps to determine whether a code change introduced a regression.
Interact with the charts
All three charts support the same interactions:
| Action | How |
|---|---|
| View details | Move the pointer over a chart to view the detailed statistics |
| Zoom into a time range | Use the cursor to select a time period to view the statistics of the specified time period |