Closed lantz closed 5 years ago
@gizmoguy @anarkiwi My intent here (similar to the faucet end-to-end tests) was to check to make sure the new flows are present in each switch (and presumably if one is present then they all are.)
However, I am still seeing some sort of race condition occasionally where even though the flows seem to be present in all of the switches, the ping test is failing, which is unexpected.
So I may need to look into this a bit more.
We assume the flow tables have been updated if a set of new flows (vlan, host MACs) are reported from
ovs-ofctl
.iputils-arping
to test-dependencies.sh