[OpenSIPS-Users] fr_timeout CANCEL message sent to timed out dispatcher endpoint IP

[solarmon]

Hi,

Using opensips 2.4.x

I'm testing some call failure scenarios and this particular test is to drop
the IP connectivity to the dispatcher endpoints and see what happens when a
new call is made and before the endpoint is considered unhealthy.

The expected behaviour is that the INVITE will not get a response from the
dispatcher endpoint (since the IP connectivity is down, but the endpoint
has not become unhealthy yet) and will try the next dispatcher endpoint.

This failover seems to be working, but I am seeing a CANCEL being sent
(with Reason: SIP;cause=480;text="NO_ANSWER"), which I assume is because
the fr_timeout (5 seconds) counter has been exceeded.

However, the CANCEL is being sent to the dispatcher endpoint IP address
that was not reachable in the first place. Why would the CANCEL be sent to
the dispatcher endpoint that did not respond to the INVITE in the first
place?

Note: I'm seeing the SIP traces as capture by HEP/Homer in this
environment, if that makes any difference.

Thank you.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.opensips.org/pipermail/users/attachments/20210629/e235179c/attachment.html>