Best Practice | API Request Throttling

Introduction

CMiC’s Cloud environment (ATLAS and NOVA) have safeguards in place to ensure that all our customers and integration partners can rely on consistent performance when calling our API endpoints. One of these safeguards throttles the number of API requests that can be made to our server.

• Our servers will process up to 15 API requests per second from a single IP address.
  
  
• Up to 100 API requests over this limit will be queued and processed in sequence. Beyond this buffer, clients will receive a 429 error indicating that too many requests have been made to this endpoint.
  
  
• Within one second, the throttle will limit the request evenly. For example, if we assume 10 requests are allowed per second:
  
  

  1 Second in Milliseconds	Requests Allowed
  100	1
  200	1
  300	1
  400	1
  500	1
  600	1
  700	1
  800	1
  900	1
  1000	1

Recommendations

• Queue API requests so the client/application is never exceeding 15 requests per second.
  
  
• Add a buffer of 100ms per request to further guarantee that your requests will stay within our limits.
  
  
• Handle a 429 error within your integration code.