Hello,
I want to know on when will the rate limit of jira rest api exceeds.
for example we have 10000 users using our app, how many requests can be made in 1 second or minute, how does this rate limiting exactly works ?
Thanks
What @Ismael Jimoh mentioned is valid for Jira, here is the corresponding Confluence developer documentation: https://developer.atlassian.com/cloud/confluence/rate-limiting/
Theory: As you will see, it will return HTTP response headers with hints.
Reality: In our experience, certain REST API end-points have no rate limits, they simply fail when overloaded. Therefore I wouldn't be surprised if you don't receive those header, but your request doesn't succeed with 200...
Hi @Sunilkumar k
You can learn more about api limits here: https://developer.atlassian.com/cloud/jira/platform/rate-limiting/
I am hoping it answers your question.
Hi @Ismael Jimoh in the document it is mentioned that the "X-RateLimit-Limit ", will be there in the headers of the response, but could'nt see that, can you please help me by saying that from where i can get X-RateLimit .
This will be returned when you make your call. As mentioned in the document, if you get a 429 response, typically, you should see the x-ratelimit-limit.
Also note what @Aron Gombas _Midori_ mentioned. I haven’t run into many issues where I exceeded this but if this isn’t thrown all the time then it’ll be hard to say.
Hello @Aron Gombas _Midori_
Are you able to advise of the specific Confluence endpoints that fail 'silently' and don't return a 429 response code if the rate limit is reached?
429
It's good coding practice to check for 500 series errors in responses, as they often indicate you've reached a resource limit, which can easily be mistaken as a 'failure':
500
Some transient 5XX errors are accompanied by a Retry-After header. For example, a 503 response may be returned when a resource limit is reached. While these are not rate limit responses, they can be handled with similar logic as outlined below.
5XX
Retry-After
503
My latest memory is that the content property "write" can lead to problems if you bulk-write content properties to a large number of contents in a short time window.
The REST API end-point returns with success or maybe with sporadic errors, but the database powering Confluence may have difficulties with handling the load. The expectation would be that rate limiting rejects the requests if they are overwhelming Confluence. (Or, if the rate limiter doesn't reject them, then I would assume that they could be handled without any side-effect.)
It looks like you're new here. Sign in or register to get started.
