MCLAG DELL POWERSWITCH E3248P-ON

Hello Ladies & Gentlemen.

I am facing a strange scenario whereby I believe it can be attributed to a STP issue on the Dell PowerSwitch E3248P-ON switches.

Brief on my current network setup

I do have a fortigate device in standalone mode; all my VLAN gateways have been defined on an aggregate interface. This particular aggregate interface connects to both of the switches on port Eth 1/47 on CSWA and port Eth 1/47 on CSWB. These particular interfaces are part of a portchannel. (NOTE THAT I AM NOT USING ANY FORM OF NEGOTIATION PROTOCOL BETWEEN THESE DEVICES.) I do have two downstream switches (S4112T) that i have configured VLT on them for redundancy purposes; a port channel has been configured to connect these two S2114T devices to form a full mesh design. The downstream portchannel is part of my mclag domain.

My mclag peer link is a L3 portchannel with the source ip being the ip configured on the CSWA's port channel and peer ip the other device's L3 portchannel's IP address.

Issue Description

 

Edge devices connecting to the downstream S4112T switches and E3248P-ON behave as if there is a loop on the network; frequent packet drops when pinging the gateway. When I remove one of the upstream connections to the fortigate; the switch remaining with the uplink connection continues to pass traffic without any anomaly..

Has anyone faced this and what was the workaround ? Would you recommend a L2 portchannel as the mclag peer link or a L3 ?

Kindly assist