API Requests Hanging Intermittently - No Response from REST and GraphQL Endpoint

[Muwahid Ahmed]

Note: While this ticket relates to Jira, the issue is specifically with API connectivity (both REST API and GraphQL API) rather than the Jira interface itself. I'm unsure where to put this specifically so let me know if this should be posted somewhere else.

I'm experiencing intermittent issues with API requests to our Atlassian Cloud instance  where some requests hang indefinitely without returning any response or error.

For example, if I send 

curl https://{ORG}.atlassian.net/rest/api/3/myself \
-u "{USER}:${ATL_TOKEN}" \
-H "Accept: application/json"

it may hang, and the whole day it may be like this where I will send the request and it receives no response, not even an error. Though strangely enough, other days this will be fine and return a response, and it will be fine the whole day. 

Issue Details:

• Affected Endpoints:
  • REST API: https://{org}.atlassian.net/rest/api/3/myself
  • GraphQL API: (general endpoint)
• Behaviour:
  • Requests hang indefinitely with no response, no timeout, and no error message
  • The issue is intermittent - some days the API works fine, other days it fails completely
  • Only seems to occur on authenticated requests. For example, if I send a request to `https://{org}.atlassian.net/rest/api/3/serverInfo` then this request returns a response fine.
  • This has been occurring over the past few days since I started using the API. 
• Authentication Method:
  • Using username and API token via Basic Authentication
• Impact:
  • This is primarily affecting our application that relies on the GraphQL API
  • The inconsistent behaviour makes it difficult to troubleshoot or work around

What I've Tried:

• Verified credentials are correct (when it works, authentication is successful)
• Tested on different days - behaviour is inconsistent

[Ayesha Sohail]

Since this only happens with authenticated requests and works fine on some days, it could be related to auth handling, rate limits, or something backend/network-related. The fact that unauthenticated endpoints respond normally is a helpful clue.

You might try regenerating your API token, adding timeouts, or testing from a different network to narrow it down.