Claudio Jeker wrote:
On Wed, Aug 01, 2007 at 08:00:26AM -0500, Jacob Yocom-Piatt wrote:
here is a quick clarification of the cabling
switch 1--------A-------switch 2
\--------------B----------/
where making connection B caused the drama. switch 1 is a nice managed
switch and switch 2 is a random POS.
is this the expected behavior with carp when a cabling topology like
this comes up sans trunking, etc? clues about what happened would be great.
You created a loop in the L2 broadcast domain and because carp is using
multicast for the hellos you had a nice network meltdown (maybe your
managed switch was realizing that and blocked the traffic).
timo, i don't have STP enabled. dunno much about it besides that i don't
use it :).
ah, thanks for clearing that up, claudio. that was my guess: the carp
multicasts got looped.
cheers,
jake
