[rsyslog] Question on the tls stalling issue

[Rory Toma via rsyslog]

Any progress? We 'd like to switch to relp (compressed, encrypted) but I
imagine I'll have the same issue, and I can't easily generate the
traffic needed to test. We have a work around for the regular issue.
Each syslog machine runs two rsyslog processes on a different port
(non-TLS). A local nginx instance takes the SSL traffic. terminates it,
and least_conn connects the data to the two syslog processes. This has
been working well for weeks.

This solution does not work for the relp case, as it appears with the
compression, it does not work to blindly decrypt, the nginx proxy would
have to understand the payload, which it does not.