A mangrove-creek site is instrumented and looks green in Application Insights while the public homepage is unreachable from Europe. What complementary check should the developer add?
Select an answer to reveal the explanation.
Short Explanation
SDK telemetry is inside-out—it only sees work that already hit the app. Availability tests are outside-in probes. Use both; do not swap in Redis.
Full Explanation
In-process Application Insights SDK telemetry is inside-out: it only sees work that already reached the process. The mangrove-creek site can look green while Europe cannot reach the public homepage if clients never hit the app. The complementary check is a Standard availability test—an outside-in probe from Azure test locations that fails when the URL is unreachable even if SDK charts stay quiet.
Doing nothing because in-process SDK metrics always prove worldwide reachability fails the core concept: green request charts can mean “nobody got through,” not “everyone is fine.”
Azure Cache for Redis as a substitute for outside-in probes fails because Redis health does not prove public homepage reachability from Europe.
Disabling the SDK so green charts cannot mask an outage fails because it removes useful inside-out diagnostics; the fix is to add availability tests alongside the SDK, not to blind the app.
Exam caveat: use both layers—SDK for failures and dependencies after traffic arrives, availability tests for client reachability—especially behind Front Door, App Gateway, or regional DNS. Operational check: block or break the public URL in a non-production slot, confirm SDK charts stay calm while the availability test fails, then restore and verify both go green.