Announcing the Saviynt Knowledge Exchange unifying the Saviynt forums, documentation, training,
and more in a single search tool across platforms. Read the announcement here.

API rate limits - v23.9

ksadhashiva
New Contributor
New Contributor

Hello,

Lately we are observing Saviynt API throwing exceptions on rate limits post upgrade to v23.9 from v23.6, could you please share the below details for Saviynt APIs.

status: 429, body: <html> <head><title>429 Too Many Requests</title></head> <body> <center><h1>429 Too Many Requests</h1></center> <hr><center>nginx</center> </body> </html>
  • X-RateLimit-Limit: The maximum number of requests allowed within a specific time frame.
  • X-RateLimit-Remaining: The number of requests remaining in the current time frame.
  • X-RateLimit-Reset: The timestamp when the rate limit will reset.

Thanks,

Karthik

9 REPLIES 9

Dhruv_S
Saviynt Employee
Saviynt Employee

Hi @ksadhashiva 

Please refer to the below forum post for API Rate limit for Saviynt APIs.

Saviynt API Rate Limiting & Performance Impact Ass... - Saviynt Forums - 24025

Regards,
Dhruv Sharma
If this reply answered your question, please accept it as Solution to help others who may have a similar problem.

ksadhashiva
New Contributor
New Contributor

Thanks for your reply.

I think the details mentioned in the reference link, is more related to network calls and not specific to API requests.  We have 2 instances DEV & PROD. DEV is updated to v23.9 where as PROD is still on v23.6. rate limiting issue was observed on DEV and not on PROD.

so the issue is more pointing towards limits imposed on API with latest versions. So if we can get the API rate limits details then we can configure our clients to handle.

Hi @ksadhashiva , we will check and let you know. Could you please confirm if it is happening for some specific applications or all of them. Some of configuration related to API limits are handled connectors and they are specific to the connections.

Regards,

Dhruv Sharma

Hello,

Thanks for the reply, this is not specific to an application rather is more specific to Saviynt APIs.

So it would be nice, if we can get rate limit stats.

 

Thanks,

Karthik

 

nimitdave
Saviynt Employee
Saviynt Employee

@ksadhashiva , just checked an found below information on new rate limit details.

As part of 23.9, EIC has implemented an API rate-limiting feature. Here are the details:

Global Rate Limit -> Total of 10,0000 requests per sec is allowed. If this is exceeded, you should get HTTP error code 429. Local Rate Limit -> 2000 requests per minute or 200 per second per IP .If this is exceeded, you should get HTTP error code 429.

Please confirm is your application is invoking APIs with in this limits.

@nimitdave : Was this mentioned anywhere in document? I was going through release notes, what's new and API documentation no where I found this information. Can you please share the respective documentation details if it is already documented? If not can you provide feedback to update the same in respective documentation? 

Whats-New 

Release-Notes-23-9 

API Documentation-v23.9 

Also if we see such errors how do we overcome? is there any wait time we can pass? If so can you please share those details?


Regards,
Saathvik
If this reply answered your question, please Accept As Solution and give Kudos to help others facing similar issue.

Is this information documented anywhere which we can refer?

Also , can you please highlight more details on Global rate limit and Local Rate limit?

Regards
Manish Kumar
If the response answered your query, please Accept As Solution and Kudos
.

Thank you for your reply!

Does this limit applies to UI calls also? Our team is getting 429 responses not only in API but also in UI.

We will test the 200 rps limit per IP based on your information.

It would also be great to document rate limiting in general, so we can refer to the documentation during the development.

 

 

nimitdave
Saviynt Employee
Saviynt Employee

@oschelkunov , yes this applies to UI calls also as the UI also invoke REST APIs internally.

We are getting this documented.