My Azure application gateway often responds with 504 / 502

Question

My Azure application gateway often responds with 504 / 502

cargobuddy 20

Not sure, but my azure application gateway often results with 504 or 502. The issue is intermittent. For example, this is the graph of the last 4 hours:

User's image

What can be the root cause of this issue?

Vallepu Venkateswarlu 10,430 Reputation points Microsoft External Staff Moderator

2026-06-16T14:13:31.2833333+00:00
Hi cargobuddy,

Welcome to Microsoft Q&A Platform

The Intermittent 502 / 504 responses from Azure Application Gateway are commonly tied to backend response timeouts or backend connectivity/health issues. Based on the docs, the most likely root cause is the backend taking longer than the timeout configured in the backend HTTP settings.

504 Gateway Timeout (most common with backend timeouts)

For Application Gateway v2, the gateway returns 504 when the backend response time exceeds the time-out value configured in the Backend HTTP settings.

So if you see spikes at specific times, it often points to the backend being slower during those windows (processing delays, dependencies, etc.).

Check the Backend HTTP settings timeout and compare it to the backend’s real response time during the incident window.

From the docs: if the backend can’t reliably complete within the timeout, either:

optimize the backend to respond faster, or

increase the HTTP backend timeout in the gateway settings.

502 Bad Gateway (often related to request timeout or backend connectivity)

For 502, the gateway can also hit a request time-out / connectivity issue with the backend (and in v1 this can surface as 502). Other common 502 causes include:

NSG / UDR / custom DNS blocking access to backend pool members

backend instances not responding to health probes

backend address pool is empty

all backend instances are unhealthy or not actually serving the app

You can verify:

Verify Backend health (make sure there are healthy instances and the probe configuration is correct).

Confirm the backend address pool is not empty.

Check that NSG/UDR/custom DNS aren’t preventing gateway-to-backend traffic.

How your timing graph fits Your screenshot shows repeated “failed requests” bursts. Per the documentation, that pattern matches situations where:

backend latency occasionally exceeds the configured timeout (leading to 504), and/or

backend connectivity/health becomes unreliable for brief intervals (leading to 502).

References:

504 / 502 troubleshooting and timeout behavior (backend response exceeds configured timeout):

Troubleshoot 502 request timeout / connectivity issues:

Troubleshoot 502 causes (backend health, NSG/UDR/DNS, empty backend pool):
Troubleshoot 504 Gateway Timeout errors:
Resolve 5xx/504 errors returned by Application Gateway:

Please "upvote" if the information helped you. This will help us and others in the community as well.
cargobuddy 20 Reputation points

2026-06-16T18:12:56.4533333+00:00

Please mark this issue as resolved. My timeout was default, which was too low for our requirements.
Vallepu Venkateswarlu 10,430 Reputation points Microsoft External Staff Moderator

2026-06-17T08:20:52.4966667+00:00

Hi cargobuddy,

Thank you for the update.

I'm glad to hear that the issue has been resolved. Based on your findings, the root cause was that the configured timeout value was too low for your application's requirements.

After adjusting the timeout setting appropriately, the issue was resolved successfully.

1 answer

Your answer

cargobuddy 20 Reputation points

2026-06-16T18:12:56.4533333+00:00

Please mark this issue as resolved. My timeout was default, which was too low for our requirements.
Vallepu Venkateswarlu 10,430 Reputation points Microsoft External Staff Moderator

2026-06-17T08:20:52.4966667+00:00

Hi cargobuddy,

Thank you for the update.

I'm glad to hear that the issue has been resolved. Based on your findings, the root cause was that the configured timeout value was too low for your application's requirements.

After adjusting the timeout setting appropriately, the issue was resolved successfully.

Answer 1

The confirmed root cause here was the backend request timeout. Application Gateway waits only for the timeout configured in the Backend Setting associated with the route. On v2, when the backend does not respond within that interval, the gateway returns 504; timeout or connectivity failures can also appear as 502 depending on the SKU and failure path.

Increase the request timeout to a value longer than the slowest legitimate backend response, then verify the change has finished provisioning. Also align any upstream web-server or reverse-proxy timeout so it does not expire first.

Use Application Gateway access logs and backend latency metrics to choose the value rather than setting it arbitrarily. If 502 responses continue when requests are not slow, check Backend Health, the probe host/path, DNS/NSG/UDR reachability, and the backend TLS certificate/SNI. Increasing the timeout fixes slow responses; it does not fix an unhealthy or unreachable backend.

Vallepu Venkateswarlu 10,430 Reputation points Microsoft External Staff Moderator

2026-07-15T13:50:20.6766667+00:00

Hi Christos Panagiotidis,

Thank you for your answer.

The customer has already identified the root cause and resolved the issue independently.