Wired Intelligent Edge

Hello all,

Reading the old Aruba 8320 Link Aggregation Guide for ArubaOS-CX 10.00 (Edition 1 of March_2018 or Edition 2 of April 2018) now superseded by the newer ArubaOS-CX Virtual Switching eXtension Guide for ArubaOS-CX 10.01 (Edition 1 of July 2018) I noticed that ISL troubleshooting on VSX, compared to ISL troublehooting on MCLAG, doesn't list the case of "Traffic redirect failure" (described as "When one side of the MCLAG is down, traffic is redirected through the ISL. Sometimes the redirection does not
occur." on older Aruba 8320 Link Aggregation Guide for ArubaOS-CX 10.00 guides).

So my question: considering current ArubaOS-CX VSX implementation (say ArubaOS-CX 10.01.0020 to stay on the very latest) is "Traffic redirect failure" an event that is not going to happen anymore?

If I actually check both VSX nodes (Primary and Secondary), I respectively find this:

Aruba-8320-1:/home$ sudo ovs-appctl vsx/show_global
ISL bundle configuration:
        Bundle name: lag128
        MAC learning: Disabled
        Flood block: Disabled
        Redirect traffic to: None
        Egress blocked VSXs:    lag1
        lag2
        lag3
        lag4
        lag5
        lag6
        lag7
        lag8
        lag9
        lag10
        lag11
        lag12
        lag13

If the Traffic Redirect failure is still a possible event also on VSX implementation, shouldn't I find something different than None/Disabled outputs for options like MAC learning, redirect traffic's destination and flood block state on each node (at least on the node owning the Secondary role) OR the redirect (through ISL) traffic's logic in case of VSX-LAG fault on one side changed substantially between MCLAG and VSX implementations and those options are not more relevant when dealing with VSX?

Clearly the output of actual ovs-appctl vsx/show_global shell command can't be equal to the one of (superseded) ovs-appctl -t hpe-mclagd mclag_filter_dump shell command but I saw some similarities so I asked.

This is very in depth OVSDB troubleshooting that would require lot more details to get the complete answer.

The ovs-appctl command that applied for 10.0 is command that might have been meaningul for support. I'm not sure that in 10.1 the concern still applies. If there is CLI commands option that does not offer the level of troubleshooting you need for VSX, please elaborate.