MXL configuration and TOR switching

Hi there,

I'm a network wannabe, but desperately trying to piece together some robust knowledge allied with some solid 'in the field' experiences.

My latest experience involves the wonderful world of Dell M630 blades and MXL and my question is on the breakout from the x520 dual port 10Gb on the blades. ESXi 6.5 will run on each blade, but with vSphere Standard licensing so no distributed vSwitches and therefore no NIOC. So the two ports will be teamed at the ESXi host level and run all of the virtual machine data

There are two unstacked MXL in the M1000e chassis, in A1 and A2, so each MXL hits one of the 10Gb ports on the x520. Each MXL is equipped with a FlexIO 4 port SFP module for connection to TOR switches.

Internally i see the 2 x 10Gb as an active/active connection. It is the breakout from the MXL to the TOR I am trying to establish. I understand that the TOR switches are also not stacked

There is:
- one 40Gb cable from A1 going to LAN network switch 1
- one 40Gb cable from A2 going to LAN network switch 2
- one 10Gb fiber connection from A1 going to Mgmt network switch 1
- one 10Gb fiber connection from A2 going to Mgmt network switch 1
- one 10Gb fiber connection from A1 going to Corp network switch

So:
For the LAN network we have 40Gb uplinks from two different, unstacked MXL going into two different TOR switches.
For the Mgmt network we have two 10Gb uplinks from two different unstacked MXL going into the same physical TOR switch.
For the Corp network we have one 10Gb uplink going to one TOR switch

My questions are around resilence and whether, because we have teamed the ESXi ports, that, with the LAN network, the TOR switches are able to receive traffic from each MXL and route accordingly to the destination. The problem i see is that if one TOR switch or one MXL went down then destinations behind that switch would be unavailable (but all ESXi resources would be available as the teaming would be able to route via the other TOR switch/MXL). Would this equate to a LAG?

Regarding the Mgmt network, the two uplinks from MXL A1 and A2 go to the same switch so my belief is this would simply need to be aggregated at the TOR level?

Many thanks for reading - if you got this far