Dear Gentle folk,

We are quite bright, but Mikrotic routers are brighter than we:)

Trying to understand why ping -R doesn't work properly.

ping works properly between all machines in my local network over which I have total control.

ping -R only works between some.

I have the following. romance and fvr are two simple linux boxes running fedora core.

MT1 and MT2 are identical RB1100 AHx2 running 6.38.3

. romance -> MT1 -> MT2 -> fvr

1.) ALL interfaces are bridged together into one broadcast domain.

2.) the link beween MT1 and MT2 is BONDED with two ports, and bonding1 is included in the bridges on both ends of MT1 and MT2.

3.) There are NO firewalls of any kind on any of the routers.

4.) All machines are one hop from each other, including those at either end that can be pinged or not, as they are all
in the same bridge group or connected to ports which are.

5.) MT1 and MT2 are running OSPF with each other as neighbors, but romance and fvr are not.

Romance is capable of ping -R any machine in its local bridged network including MT1, but can not ping -R MT2 or fvr. nor anything on the further side of fvr connected to the same bridged group through other simple switches.

fvr is NOT capable of ping -R anything, including MT2, nor any other machines connected to the MT2 bridge.

Is there something about a bonded link that would be playing havoc with ping -R?

Is there some other setting I am missing that might cause a ping -R to fail?

Thanks in advance.

Homer W. Smith
CEO Lightlink Internet

Reading your description.

1.) ALL interfaces are bridged together into one broadcast domain.

2.) the link beween MT1 and MT2 is BONDED with two ports, and bonding1 is included in the bridges on both ends of MT1 and MT2.

4.) All machines are one hop from each other, including those at either end that can be pinged or not, as they are all
in the same bridge group or connected to ports which are.

1, ok not so bad. A single broadcast domain on either MT1 or MT2. I'm tracking.
2. Hmmmmm. You created a bond and added it to the bridges on both sides. That means your broadcast domain is now stretched across the 2 devices.
4. Not sure how they can be 1 hop from one another when they are in the same broadcast domain.

Please clarify some configuration dumps. In particular the interface list, bridge port list and any VLANs you are doing. Also a drawing with you are trying to do is always helpful.

This problem has been resolved.

If use ip-firewalls is set to yes in the bridge menu, ping -R packets will be dropped and not sent on and not
replied to, even if the pinging host is pinging the mik directly.

Once use-ip-firewalls is turned off and fast bridging is allowed again, ping -R works brilliantly.

Thanks for all who read my original post.

Homer W. Smith
CEO Lightlink Internet