Loop issue multiple lag across M4300 stack

Question

Hi,I'm experimenting different scenarios with 2 M4300 24X and 4 S3300. I'm trying the configuration below, but when I connect the 8 port between the S3300 and the M4300 in the LAG I experience networks issues, compatibile with a loop problem. Even with 4 port in the lag the loop occurs.&nbsp;Is this configuration supported?&nbsp;&nbsp;&nbsp;

JohnC_V · Answer

andrea_g,

&nbsp;

Welcome to our community! :)

&nbsp;

It seems that you are having issues creating a LAG on your switches. Have you tried creating a LAG from 1 S3300 switch to 1 M4300 just to test if that will work before adding all the switches on your network? Please also check if STP on all switches is enabled.

&nbsp;

Regards,

&nbsp;

John

NETGEAR Community Team

andrea_g · Answer

Enabling the&nbsp;Local Preference Mode in the lag settings solved this issue.&nbsp;The Local Preference Mode setting is not explained in the manual, but in the S3300 manual there are more info:&nbsp;M4300 Manual:Use Local Preference Mode to Enable or Disable the LAG interface’s local preference mode.The default is Disable.S3300 ManualLocal Preference Mode. Enable or disable the LAG interface’s Local Preference&nbsp;mode. Local preference is one of the properties of a LAG interface which is intended&nbsp;for a Stacking environment. This is useful when the LAG is formed with ports from&nbsp;across the units. In such a scenario, when this feature is enabled, any known unicast traffic sent to the LAG uses only the LAG interface on the local unit. This ensures that the known unicast traffic, destined to the LAG, does not cross the external stack link when the LAG has a member or members on the local unit. Local preference does not impact behavior with respect to unknown unicast, broadcast and multicast traffic

bbs2web · Answer

I would suggest that you cable the switches up whereby each leaf connects to each spine redundantly, that would give you 4 x 10 Gbps capacity from one leaf to another. Then distribute your LAGs redundantly over the leaf switches and turn on local preference mode.

To clarify, I believe the following reference diagram in the Netgear documentation to be un-optimal. We run this in production on a number of stacks without issues for 4+ years:

My understanding of the leaf-spine (Spine & Leaf) CLOS network design is to ensure each point on the spine and leaf switches has multiple paths to another in the same layer. Spine switches should connect to each leaf, never another spine. Each leaf should also never connect to another leaf, only to the spine switches.

We implemented a stack of 3 x M4300-8F8X (spine) with 4 x M4300-96X (leaf) switches.We setup distributed LAGs from the X96 leaf switches to servers and enabled local preference mode. This results in the 96X forwarding received traffic via the destination's LAG port on the same switch traffic arrived on and stops traffic unnecessarily loading stack uplinks. This works as expected, although Netgear documentation is extremely poor (something along the lines of "enabling local preference mode enables local preference").

If each 96X has 6 x 10G stack members (2 to each 8X8F) there should be 6 equal distance hops between each 96X leaf switch. If server A sends data to server B and both only have 2 x uplinks running LACP (each uplink in a unique 96X) a single steam of data would have 6 possible choices.