Has anyone else who is running EIGRP Over the Top come across this bug? I am really liking OTP, however I find the LISP component of it very unreliable when it comes to the LISP interface itself. It seems like LISP just randomly picks another interface on the router to use as it's unnumbered IP. This IP shows up in trace routes as it's the outbound interface for the next-hop presented by LISP. This is very confusing for our clients as they randomly see a 192.168.1.2 in their traces, and it's hard for me to explain why.
It seems like EIGRP OTP LISP should always use the WAN interface of the SP as it's unnumbered IP/interface. SOMETIMES this does happen, but most of the time it's a different interface, or the internal Linux interface on IOS XE per bug ID CSCun72490.
I'm less than impressed by the lack of updates on EIGRP OTP.