Intermittent “Connection Failed with IP Address” Error in Azure API Management

Question

Intermittent “Connection Failed with IP Address” Error in Azure API Management

Vijay Sutaria 0

Hello,

Over the past few days, we’ve observed that our Azure API Management (APIM) instance intermittently throws an error for a few minutes and then resumes normal operation without any intervention.

Upon investigation, I found the following error in the logs (End-to-end transaction details):

“Connection failed with IP address”

Interestingly, after a few minutes, when I ping the same IP address, it responds as expected.

I’m trying to determine which resource is associated with that IP address, but haven’t been able to identify it yet.

Could you please help me understand:

How can I identify which Azure resource is linked to that IP address?

What could be the potential root cause of this intermittent connectivity issue?

What’s the best approach or recommended solution to prevent or troubleshoot this type of issue?

Thank you for your assistance!

Best regards,

Vijay

Anurag Rohikar 3,185 Reputation points Microsoft External Staff Moderator

2025-11-10T17:50:23.8066667+00:00
Hello Vijay Sutaria,

Intermittent “Connection failed with IP address” errors in Azure API Management (APIM) typically indicate transient network connectivity issues between APIM and the backend service. Since the connection later resumes without any manual intervention, this often points to short-lived network routing, DNS, or backend availability issues.

Identifying the Resource Linked to the IP Address:

To find out which Azure resource is associated with the IP address you're seeing in the error logs, you can follow these steps:

Azure Portal: Check the "Overview" blade of your APIM instance to view the public Virtual IP (VIP).

Check NSG Rules: If the APIM is deployed within a virtual network, ensure that your Network Security Groups (NSGs) are configured correctly. The NSG rules should not block inbound traffic to the necessary ports.

Possible Causes and Troubleshooting Steps:

1. Verify APIM Networking Configuration

If your APIM instance is integrated with a Virtual Network, verify the following:

DNS resolution: Ensure the DNS servers configured for APIM can resolve backend endpoints consistently.

NSG/UDR rules: Check for any network security group or user-defined route that could intermittently block outbound connectivity.

Firewall or NVA: If there’s a network virtual appliance or firewall in the path, confirm no session timeout or packet inspection policy is causing temporary drops. Reference: Troubleshoot networking issues in Azure API Management

2. Check Backend Availability

If the backend service is hosted in Azure (App Service, AKS, VM, etc.), verify whether there were any scaling or restart events around the time of the failures. During scale or redeploy operations, the backend IP address or routing can temporarily change, leading to intermittent connection errors.

3. Correlate Logs and Monitor Connectivity

Review APIM diagnostic logs (under Monitoring > Diagnostic Settings) to check if the failures correspond to DNS or timeout errors.

Enable Connection Monitor or Azure Monitor Logs to correlate the failed timestamps with backend activity or platform maintenance events. Reference: Diagnose connectivity using Connection Monitor

To help narrow down the issue further, please confirm the following:

Is your APIM deployed in internal or external VNET mode?

What type of backend service are you calling (App Service, AKS, VM, or external API)?

Is the failing IP address public or private?

Do you observe this issue for a specific API or multiple endpoints?

Are there any custom DNS or firewall configurations in your environment?

Were there any recent scaling or network configuration changes when this behavior started?

If after verifying the above points you still experience the same issue, please share the requested details along with a sample timestamp and correlation ID. This will help in analyzing logs and identifying the exact cause.

Hope this helps.

Additional References:

Troubleshoot networking issues in Azure API Management

Azure Network Watcher IP Flow Verify

IP Address of your API Management Service

Troubleshooting client response time-outs and errors with API Management
Anurag Rohikar 3,185 Reputation points Microsoft External Staff Moderator

2025-11-11T07:52:13.0033333+00:00

Hello Vijay Sutaria, following up to see if you have a chance to check my previous response and helped, do let me know if you have any further questions on this. Thank-you!
Vijay Sutaria 0 Reputation points

2025-11-11T09:38:19.3966667+00:00

Hi Anurag,

Thank you for getting back to me. I appreciate your time and follow-up on my earlier query. Please find my responses to your questions below:

1. Is your APIM deployed in internal or external VNET mode? No VNET is currently configured.

2. What type of backend service are you calling (App Service, AKS, VM, or external API)? The backend service is an Azure Function App.

3. Is the failing IP address public or private? The IP address is auto-generated by Azure.

4. Do you observe this issue for a specific API or multiple endpoints? The issue occurs across multiple endpoints, specifically for all Azure Function Apps that are routed through Azure APIM.

5. Are there any custom DNS or firewall configurations in your environment? No, there are no custom DNS or firewall configurations in place.

6. Were there any recent scaling or network configuration changes when this behavior started? No, there have been no recent scaling or network configuration changes.

Thanks
Anurag Rohikar 3,185 Reputation points Microsoft External Staff Moderator

2025-11-13T05:51:25.4966667+00:00

Hello Vijay Sutaria, just checking to see if you have a chance to check my previous response and helped, do let me know if you have any further questions on this. Thank you!
Anurag Rohikar 3,185 Reputation points Microsoft External Staff Moderator

2025-11-14T09:58:57.2433333+00:00

Hello Vijay Sutaria, following up to see if you have a chance to check my previous response and helped, do let me know if you have any further questions on this. Thank you!
Vijay Sutaria 0 Reputation points

2025-11-17T09:13:41.93+00:00

Hi Anurag, The details you have provided is too generic and doesn't taking me to the actual solution. Good thing is, luckily we haven't seen this issue again any more after I posted this question and hopefully it will not happen again. but if ever it will happen again I am still not sure how this will be fixed.
Anurag Rohikar 3,185 Reputation points Microsoft External Staff Moderator

2025-11-18T08:27:57.8+00:00

Hello Vijay Sutaria, I have reached to you via private message for further update.
Vijay Sutaria 0 Reputation points

2025-12-23T09:29:14.8033333+00:00

@Anurag Rohikar issue is still there, it starts suddenly for few minutes anytime/any day for few minutes and start working fine itself. I have sent you details in personal message
Anurag Rohikar 3,185 Reputation points Microsoft External Staff Moderator

2025-12-24T01:19:33.9933333+00:00

@Vijay Sutaria , I have reached to you via Private message, could you please provide us with the required information. Thanks for your patience on this.

1 answer

Your answer

Anurag Rohikar 3,185 Reputation points Microsoft External Staff Moderator

2025-11-11T07:52:13.0033333+00:00

Hello Vijay Sutaria, following up to see if you have a chance to check my previous response and helped, do let me know if you have any further questions on this. Thank-you!
Vijay Sutaria 0 Reputation points

2025-11-11T09:38:19.3966667+00:00

Hi Anurag,

Thank you for getting back to me. I appreciate your time and follow-up on my earlier query. Please find my responses to your questions below:

1. Is your APIM deployed in internal or external VNET mode? No VNET is currently configured.

2. What type of backend service are you calling (App Service, AKS, VM, or external API)? The backend service is an Azure Function App.

3. Is the failing IP address public or private? The IP address is auto-generated by Azure.

4. Do you observe this issue for a specific API or multiple endpoints? The issue occurs across multiple endpoints, specifically for all Azure Function Apps that are routed through Azure APIM.

5. Are there any custom DNS or firewall configurations in your environment? No, there are no custom DNS or firewall configurations in place.

6. Were there any recent scaling or network configuration changes when this behavior started? No, there have been no recent scaling or network configuration changes.

Thanks
Anurag Rohikar 3,185 Reputation points Microsoft External Staff Moderator

2025-11-13T05:51:25.4966667+00:00

Hello Vijay Sutaria, just checking to see if you have a chance to check my previous response and helped, do let me know if you have any further questions on this. Thank you!
Anurag Rohikar 3,185 Reputation points Microsoft External Staff Moderator

2025-11-14T09:58:57.2433333+00:00

Hello Vijay Sutaria, following up to see if you have a chance to check my previous response and helped, do let me know if you have any further questions on this. Thank you!
Vijay Sutaria 0 Reputation points

2025-11-17T09:13:41.93+00:00

Hi Anurag, The details you have provided is too generic and doesn't taking me to the actual solution. Good thing is, luckily we haven't seen this issue again any more after I posted this question and hopefully it will not happen again. but if ever it will happen again I am still not sure how this will be fixed.
Anurag Rohikar 3,185 Reputation points Microsoft External Staff Moderator

2025-11-18T08:27:57.8+00:00

Hello Vijay Sutaria, I have reached to you via private message for further update.
Vijay Sutaria 0 Reputation points

2025-12-23T09:29:14.8033333+00:00

@Anurag Rohikar issue is still there, it starts suddenly for few minutes anytime/any day for few minutes and start working fine itself. I have sent you details in personal message
Anurag Rohikar 3,185 Reputation points Microsoft External Staff Moderator

2025-12-24T01:19:33.9933333+00:00

@Vijay Sutaria , I have reached to you via Private message, could you please provide us with the required information. Thanks for your patience on this.

Answer 1

Hello Vijay Sutaria,

Thanks for confirming the details. Given there’s no VNET, custom DNS, or recent configuration changes, this behavior most likely points to short, transient connectivity issues between APIM and the Function App’s public endpoint that usually resolve on their own.

Analysis and Recommendations:

Transient Backend Connection Drops Azure Function Apps, when accessed over public endpoints, can experience short-lived connection drops due to load balancer changes, scaling operations at the platform layer, or DNS propagation delays. These are typically self-healing and may appear intermittently in APIM traces.
- You can validate backend health using the “Availability and Performance” metrics under Function App → Monitoring → Metrics → Availability.
- Check APIM diagnostic logs to confirm whether the failures are due to DNS resolution or timeout exceptions. References: Monitor Azure Functions - Metrics, Azure Status
Review APIM Timeout and Retries
- Ensure your APIM backend request timeout settings are sufficient for the Function App’s response time.
- Consider adding retry policies within your API Management policies to handle transient failures gracefully. Reference: Set retry policies in Azure API Management
Monitor Network Path and Latency
- Use Network Watcher Connection Monitor to periodically test connectivity between APIM and the Function App endpoint.
- This will help confirm if packet loss or transient failures occur at the network layer. Reference: Diagnose connectivity using Connection Monitor
Enable Application Insights Correlation
- Enable Application Insights for both APIM and the Function App to track end-to-end dependency calls.
- Correlation logs can help identify whether the failure originates from the APIM side, network, or Function App itself. Reference: Monitor and troubleshoot API Management
Platform Maintenance or Transient Faults Even without any configuration changes on your side, such transient connectivity issues may occur during underlying platform maintenance or brief backend reallocation. These usually recover automatically within sometime.

Additional Considerations:

Function Apps on Consumption plans share outbound IPs that can change due to platform events, which adds to potential instability. If feasible, consider moving to a Premium or App Service plan and integrate both APIM and Function Apps within a shared VNET to gain stable IP addresses and improved routing.
If you have high request volumes or spikes, SNAT port exhaustion can occur on APIM's public IPs, causing intermittent failures. Reducing connection churn via HTTP keep-alives or scaling APIM tier and backend may help mitigate this. References: Deploy Azure API Management Instance to External or Internal VNET, For Azure API Management SNAT port limitations and mitigation

Additional References:

Retry policies in API Management

Hope this helps clarify and let us know if you have further questions. Thank you!

Share via

Intermittent “Connection Failed with IP Address” Error in Azure API Management

1 answer

Your answer