Hi Lori, I have created a new issue in your github issue tracker as requested (issue 7).
Best Regards, Paris On 26/1/2012 14:19, Lori Jakab wrote: > Hi Paris, > > On 01/26/12 12:22, Paris Charalampou wrote: >> Hi *, >> >> When a host(hostA) with ip address tries to communicate with the EID >> of mobile node(hostB) with RLOC on the same subnet as hostA, all >> replies from mobile node are not sent to the PxTR but directly to >> nodeA, encapsulated in lisp packets. Since hostA is not in a lisp site >> nor is lisp-aware, packets are dropped. The only way to by-pass the >> issue is by removing the static route from hostB for the subnet of its >> RLOC. >> I am not sure that these packets should be encapsulated to lisp. But >> even if they are encapsulated I think that lispmob should forward them >> to PxTR (as in the workaround) and not on LAN. > Thanks for reporting this issue. > > I think the correct solution is not to remove the static route, but > instead to not encapsulate at all towards non-LISP destinations that are > on the same subnet as any of the active RLOCs of the MN. We will look > into this. > > Can you please file an issue in our Github issue tracker (no milestone): > https://github.com/LISPmob/lispmob/issues > > Regards, > -Lori > >> Paris >> >>
