T.Togashi | 2018-01-19 04:36:02 UTC | #1
Hello,
Regarding API calls, we sometimes get 429 error even though the API rate limit has not been exceeded. It also occurs in the Collaborate UI, and several users get the error at the same time.
<img src="//inin-prod-use1-developerforum.s3.amazonaws.com/original/1X/6146b3c36e0b5a6fb8f8c77fcd9764820b07e5ca.PNG" width="690" height="394">
Apart from the per-token basis API rate limits, there seems to be the per-organization basis API rate limits. Is there any documantation about this? If not, please make it avaiable.
Thank you.
tim.smith | 2018-01-19 16:40:20 UTC | #2
This is working as expected. There multiple rate limiting triggers. A well-behaved app should handle the 429 in any case; the rate limiting headers only represent the global rate limiting applied at the API layer, not other rate limiters. The internal rate limits are intentionally unpublished as they are often complex calculations and some have varying rates that change based on server load.
T.Togashi | 2018-01-22 03:57:53 UTC | #3
Thanks for your reply. If other organizations are heavily using server resources, will this affect our organization's API rate limit? Or are server resources clearly separated for each organization, are they designed to not affect other organizations?
tim.smith | 2018-01-22 20:16:49 UTC | #4
Back-end rate limits are generally only applied to some combination of per token, user, or organization.
some have varying rates that change based on server load.
I apologize, that statement was a little misleading. You will not be rate limited simply because another customer has high activity. I was referring to catastrophic situations that are not part of normal operation.
Poulain_Mathieu | 2018-02-15 00:26:20 UTC | #5
Hi Tim,
When we are experiencing Error 429, a lot of our users are getting a Purecloud Maintenance mode when trying to connect.
tim.smith, post:2, topic:2374
A well-behaved app should handle the 429 in any case; the rate limiting headers only represent the global rate limiting applied at the API layer, not other rate limiters.
In this case our custom applications are handling 429 error well, but it seems that Purecloud itself is quite suffering from those errors. Is it normal that custom API calls are impacting the normal funcitonning of Purecloud (login, answering calls...)?
tim.smith | 2018-02-15 16:22:38 UTC | #6
I can't say why you're seeing the maintenance mode screen from the OOTB app. As that's not directly related to usage of the APIs, please open a ticket with PureCloud Support to investigate that issue.
system | 2018-03-18 16:22:40 UTC | #7
This topic was automatically closed 31 days after the last reply. New replies are no longer allowed.
This post was migrated from the old Developer Forum.
ref: 2374