I have implemented 2 FRRP instances, followed all the guidelines, and for the most part, both rings work properly. Occasionally, when a "failover" or "ring down" scenario happens, not all L2 Vlans forward traffic properly, Certain VLANS dont learn mac-addresses in one direction on a particular link. is this common, and how can I prevent it? it seems to be different VLANs every time, but they are almost always newly added vlans to the member-vlan statement. new meaning they haven't been in the member-vlan statement during a failover/ring down condition before. If I remove the affected vlans from the affected link, then re-add them, MAC's are learned properly, and traffic resumes as expected.
some vlans are member vlans in both rings, but some are not as well. Last night, I had the larger ring fail, and for the first time have it also trigger the smaller ring at the exact same time in logs. these rings are essentially the figure 8 configuration in the guides. separate physical ports, different control vlans, and the switch in the middle of the figure 8 is transit in both nodes.
jurwiller
4 Posts
3293
0
Posted March 19th, 2019 00:00