sudhakar-datta

It would be amazing if there are plans for making into a particular release and timeframe. This feature will be one of the most useful feature in a production workflow.

Thanks so much for getting back so quick. Regards! Let me do that testing today. BTW, the producer is a keystone middleware (openstack) talking to keystone backend. Context is, any...

Caracal. This also happens with any ancillary service that is written to use eventlet. Let me bisect the issue as you have highlighted. It is a good idea. I will...

Thanks for pointing at the keystone changes. Looking at the diff, bit confused why a downgrade of dnspython would solve the issue. Just now tested with modified eventlet (dirty way,...

Makes total sense. Let me run through the tests by pinning crypto, and then go down the path of dnspython.

Okay, this changes breaks the eventlet - https://github.com/rthalley/dnspython/pull/908

Is this the reoccurrence of - https://github.com/eventlet/eventlet/issues/726 Also, https://github.com/eventlet/eventlet/issues/371#issuecomment-1679519870 Trace back reflects the same issue.

Fails with Python3.11.9 with Antelope as well.

> I'm a bit surprised If [rthalley/dnspython#908](https://github.com/rthalley/dnspython/pull/908) (`6c5f0c9d8086c999357531c38b831efd24b6b5ac`) is the root cause of the observed problem. > > This commit is now present in dnspython since many versions (at least...

> These TLS changes are now present in Eventlet since version 0.35.2, versions which correspond the version you highlighted in the title of this github issue: [e2c874e](https://github.com/eventlet/eventlet/commit/e2c874ee83276c1e26658fc2abcc8760ff7f835b) With dnspython >=2.4.0