Hey guys, this is not really a problem. Its me not being able to wrap my head around why something is happening the way it is.

Here we go. MPLS setup on 6 routers, all 4XXAH Routerboards all running 3.23 w/mpls-test. Linear setup with one redundant path (much like the recomended lab setup). OSPF and MPLS running just fine.

I setup a VPLS interface on the first and last routerboards and enable ospf on those interfaces (ospf running on the physical links and ospf running inside VPLS). For a few seconds this works fine, then one end (only one end, happens on both ends, but only one at a time) of the VPLS interfaces will show traffic of 600+ Mbits/s. then quiet down and then repeat. A ping is not going through when VPLS shows 600+. Is this a routing loop? I enabled loop detect on the 2 routers with VPLS, that does not help.

I then setup a TE interface on each routerboard with endpoint of the other routerboard (matching the VPLS setup), and voila everything is fine and runs smooth. The VPLS traffic behaves as it should and moves to the TE tunnel and all instabillity stops.

Long explanation, but there are to many elements for me to comprehend whats happening, but it would be really helpful if someone could point out why it is happening, since i do see a use for VPLS w/OSPF.

Regards,

Martin Madsen
Bel Air Internet

Most likely this happens because adding VPLS interface to OSPF makes it shortest path (just one hop) to remote endpoint - therefore VPLS tunnel sends data over itself, you can confirm this by monitoring route changes. TE tunnel works because VPLS tunnel prefers TE tunnel path over path according to routing table.

Hey,

Thanks, you are right. The router dead interval is 40 seconds, same as the outages and the routing table reflects the shortest path becomes the vpls interface.


Thanks Again.

Martin Madsen
Bel Air Internet