Strange Throttling Behavior on Azure API Management During High Traffic
Strange Throttling Behavior on Azure API Management During High Traffic
Hello community!!!
We are experiencing some unanticipated throttling within our Azure API Management (APIM) instance, despite our policies appearing to be set up correctly.
Here is the situation:
We have a Premium tier APIM instance that hosts several APIs.
Rate-limit policies are applied per subscription, such as allowing 1000 calls per minute.
From time to time, clients are receiving HTTP 429 (Too Many Requests) responses, even when their usage is well within the established limits.
The throttling appears to occur in bursts, frequently aligning with traffic spikes on other APIs.
Here' is what we have looked into so far:
APIM scaling and instance units are in play. Policies are applied at the product, API, and operation levels.
We are monitoring things with Azure Monitor and APIM analytics, but the 429 errors are not lining up neatly with the logs.
Network latency and client retries are also factors. We are trying to figure out if this is a policy evaluation order issue, how burst handling is working, or if there is some undocumented limit in APIM that is causing the problems.
Has anyone else seen intermittent throttling in Azure API Management when traffic seems normal?
What are the best ways to troubleshoot or visualize the actual request flow and the rate limits that are being enforced?
All Comments (5)
Hello,
Thank you for reaching out to us regarding this,
To clarify, you are encountering this issue within RDM?
Could you please specify the version of RDM you are currently using?
Would you be able to provide me with a screenshot of the error message?
Best regards,
Samuel Dery
Hello,
Thank you for reaching out to us regarding this,
To clarify, you are encountering this issue within RDM?
Could you please specify the version of RDM you are currently using?
Would you be able to provide me with a screenshot of the error message?
Best regards,
@Samuel Dery
He!!!! Thanks for getting back to me. We are mostly looking into this in the context of Azure API Management (APIM). Could you explain if RDM can be used to help figure out why APIM is throttling sometimes?
We're especially interested in:
How APIM enforces rate limits and deals with bursts, especially in Premium tier instances.
The order of policy evaluation and how it affects throttling.
The best ways to keep an eye on or see how requests are flowing and how throttling works.
We would really appreciate any advice, suggestions, or experiences you can share, as they will help us better understand and fix these kinds of problems.
Hello,
Thank you for your reply,
RDM would not be helpful with this.
Unless you are encountering this issue within RDM I'm afraid we would be fairly limited in what advice we can provide regarding this,
Best regards,
Samuel Dery
Hello,
Thank you for your reply,
RDM would not be helpful with this.
Unless you are encountering this issue within RDM I'm afraid we would be fairly limited in what advice we can provide regarding this,
Best regards,
@Samuel Dery
Hello,
Thank you for clarifying. I understand that RDM is not relevant here.
Our issue is specific to Azure API Management (APIM), and we’re trying to understand intermittent 429 throttling behavior in that context.
We’ll continue troubleshooting APIM directly, but any guidance you could offer on monitoring or analyzing API request flow and rate-limit enforcement in APIM would be appreciated if you have experience in that area.
Thanks again for your time and clarification.
Hello,
Thank you for your reply,
I see, I'm afraid this would not be something I can assist you with.
Feel free to let me know if you have questions that relate to some of our products.
Best regards,
Samuel Dery