Azure APIM Best Practices

Question

Azure APIM Best Practices

Siddappa, Santhosh Kumar 25

I wanted to share a challenge we're facing with our public-facing application that uses APIM. As it stands, we are routing UI requests to the app, but regardless of the tier, we are limited to only 100 operations within one API routing. Given that our application has over 4,000+ operations, this presents a significant limitation.

To address this, I'm looking for the best option to eliminate the need to add operations for each routing. Your insights and suggestions would be greatly appreciated.

Accepted answer

0 additional answers

Your answer

Answer 1

Sai Prabhu Naveen Parimi 2,265 Microsoft External Staff Moderator

@Siddappa, Santhosh Kumar

Summary & Generalized Answer:

You're facing a common limitation in Azure API Management (APIM), where each API is restricted to 100 operations. For applications with thousands of endpoints — especially in a public-facing, multitenant setup — this becomes a challenge.

Here are some general best practices and workarounds:

Wildcard Operations: Use a route like /api/{*path} and apply rewrite-uri policies to dynamically route requests. This avoids the need to define each operation individually.

API Segmentation: Break your APIs into logical units based on functionality (e.g., Orders, Users, Billing) or tenant-specific endpoints to stay within limits and improve maintainability.

Lightweight Policies: Keep APIM policies simple. Avoid deep conditional logic to minimize latency.

Backend Routing Logic: Shift complex routing or multitenancy logic to the backend services instead of handling it all in APIM.

Caching: Use APIM response caching or external solutions like Azure Front Door or Redis to reduce backend load and improve performance.

API Versioning: Use versioned endpoints (e.g., /v1, /v2) to organize and manage APIs more effectively.

Monitoring and Cleanup: Use Azure Monitor or APIM Analytics to identify unused operations and clean up redundant ones.

Consider Premium Tier: The Premium tier supports higher scalability, multi-region deployment, and better performance isolation for demanding workloads.

These strategies can help you maintain performance and scalability while working within APIM's operational limits.

Thank you for your kind words — I’m glad the suggestions were helpful. Please do not forget to click "Accept the answer” and “Yes” wherever the information provided helps you — this can be beneficial to other community members.

User's image

Sai Prabhu Naveen Parimi 2,265 Reputation points Microsoft External Staff Moderator

2025-05-28T06:08:26.7333333+00:00
@Siddappa,Santhosh Kumar

Thanks for bringing this up — I completely understand the issue you're facing. The 100-operation limit per API in Azure API Management (APIM) can be a real bottleneck, especially when you’ve got thousands of routes to manage.

Here are a few ways you can work around it:

Use wildcard routing: Instead of defining each operation, you can use a wildcard like /api/{*path} and route requests dynamically using policies. This lets you handle multiple routes through a single operation. Example policy:

<rewrite-uri template="/api/@(context.Request.MatchedParameters["path"])" /> <set-backend-service base-url="https://your-backend.com" />

Reference: Add a wildcard operation

Split APIs by functionality: Group operations into separate APIs (e.g., User API, Orders API, etc.) so you stay within the 100-operation limit per API. Makes things easier to manage too. Reference: API design guidelines

Use API versioning: If you’re not doing it already, versioning your APIs (like /v1, /v2) helps organize your routes and spread them across separate definitions. Reference: Publish versions of your APIs

Offload logic to policies: You can handle a lot using policies — like header injection, conditional routing, or transformation — without needing extra operations. Reference: Policy reference guide

Monitor and clean up unused endpoints: Use Azure Monitor or APIM’s built-in analytics to see which endpoints are actually being used. You might find you can deprecate or combine many of them.

Know the limits: Here's the official APIM limits document so you have full context on quotas: APIM quotas and limitations
Siddappa, Santhosh Kumar 25 Reputation points

2025-05-28T08:02:22.9333333+00:00

Thank you @Sai Prabhu Naveen Parimi . While it's great that we have this tool at our disposal, I am a bit concerned about potential performance implications that might arise down the line. Given that our application is multitenancy, it's crucial for us to maintain optimal performance across all tenants. If we do encounter performance issues stemming from the use of this wild card option, do we have alternative solutions or best practices to overcome them? Your insights and recommendations would be highly appreciated as we strive to ensure the smooth operation of our application for all users.
Sai Prabhu Naveen Parimi 2,265 Reputation points Microsoft External Staff Moderator

2025-05-28T08:42:11.28+00:00

@Siddappa,Santhosh Kumar

Thank you for raising this concern. You’re absolutely right — in a multitenant architecture, maintaining consistent performance is critical, and it’s important to be mindful of the trade-offs that come with using wildcard operations in Azure API Management (APIM).

Here are some general best practices and alternative approaches that can help address potential performance implications:

Keep policies lightweight: Complex logic in rewrite-uri or set-backend-service policies can increase request processing time. Keeping policies minimal helps maintain responsiveness. (Reference: Policy guide already shared)

Avoid deep conditional logic: Reducing nested choose or if statements within policies helps streamline execution and reduce latency. (Reference: Policy expressions link previously shared)

Segment APIs by tenant or functionality: Structuring APIs by service type or tenant (e.g., /tenant1/api/{*path}) can help distribute load and improve clarity. This approach aligns with the API design guidance already provided.

Offload routing logic to backend services: Moving detailed routing or tenant logic to the backend can help simplify and speed up processing at the APIM layer.

Implement caching strategies: Using response caching in APIM or backend systems like Azure Front Door or Redis can reduce repeated load. This was covered in the caching documentation shared earlier.

Apply API versioning: Versioned endpoints can isolate changes and simplify maintenance. (See previously shared versioning reference)

Be aware of APIM limitations: Understanding quotas and service limits ensures your design remains scalable. (Limits documentation already shared)

Scale with the Premium tier if needed: In high-load or strict SLA scenarios, the Premium tier supports scaling out to multiple regions or dedicated instances. This helps improve performance isolation and availability in multitenant deployments. ref : https://azure.microsoft.com/en-us/pricing/details/api-management/

These strategies can help balance flexibility and performance.
Sai Prabhu Naveen Parimi 2,265 Reputation points Microsoft External Staff Moderator

2025-05-30T06:08:54.7166667+00:00

@Siddappa,Santhosh Kumar

Just checking in to see if the above information was helpful. Let me know if you have any questions or need further clarification.
Siddappa, Santhosh Kumar 25 Reputation points

2025-05-30T06:26:52.7466667+00:00

Yes, they were really helpful, and I truly appreciate the effort you put into providing them.

I will begin working on implementing on your suggestions and will reach out if I need any further assistance or clarification.

Share via

Azure APIM Best Practices

0 additional answers

Your answer