Apologoes for the typo in the question title: MLAG
I want to create a core with an A-Side / B-Side set of switches that extend from our Core rack to remote data center rows that also use a A-Side / B-Side configuration for redundancy. I havee been researching MLAG to connect these rows, but all the examples I have found show a pair of upstream switches using MLAG to connect to a single downstream device/stack. Is it possible to have similar stacks in the data center row?
Each of the colored set of lines are an MLAG from each of the Core rack switches, that terminate into the data center row switch stacks.
You may notice that I expect use MLAG on the lower switches as well to accomodate 1Gig connectivity within the row. Can I use MLAG on the data center row switches (lower stacks) in the same way as the Core switches (upper stacks) to have a redundant upstream connection, or am I limited to a LAG on the lower switch pairs (and subsequently single device or stack)?
My initial design to use a firewall for access control and NAT/PAT into each of the data center rows would work well with the MLAG concept. Core switches would use MLAG to connect the firewalls and data center rows would do the same to the firewall for access..
In this case I am confident that the MLAG from the upper switches to the firewalls would work as expected, and the MLAG on the data center row switches would work for both the firewalls and the 1G connectivity below.
I am looking to have a hybrid network with both the Layer 3 routed services available through the firewalls using the design above, and layer 2 services provided via a few VLAN's on the "layer 2" design at the top to accomodate high speed transfers to shared resources that the firewall cannot facilitate due to ASIC limitations..
I think my catch is whether or not I can use 2 stacks in the data center rows, or if I have to have an MLAG from the Core to each stack in each data center row.
I want to create a core with an A-Side / B-Side set of switches that extend from our Core rack to remote data center rows that also use a A-Side / B-Side configuration for redundancy. I havee been researching MLAG to connect these rows, but all the examples I have found show a pair of upstream switches using MLAG to connect to a single downstream device/stack. Is it possible to have similar stacks in the data center row?
Each of the colored set of lines are an MLAG from each of the Core rack switches, that terminate into the data center row switch stacks.
You may notice that I expect use MLAG on the lower switches as well to accomodate 1Gig connectivity within the row. Can I use MLAG on the data center row switches (lower stacks) in the same way as the Core switches (upper stacks) to have a redundant upstream connection, or am I limited to a LAG on the lower switch pairs (and subsequently single device or stack)?
My initial design to use a firewall for access control and NAT/PAT into each of the data center rows would work well with the MLAG concept. Core switches would use MLAG to connect the firewalls and data center rows would do the same to the firewall for access..
In this case I am confident that the MLAG from the upper switches to the firewalls would work as expected, and the MLAG on the data center row switches would work for both the firewalls and the 1G connectivity below.
I am looking to have a hybrid network with both the Layer 3 routed services available through the firewalls using the design above, and layer 2 services provided via a few VLAN's on the "layer 2" design at the top to accomodate high speed transfers to shared resources that the firewall cannot facilitate due to ASIC limitations..
I think my catch is whether or not I can use 2 stacks in the data center rows, or if I have to have an MLAG from the Core to each stack in each data center row.
