C9300-M port-channel suspended due to storm control native IOS-XE *workaround provided*

[GIdenJoe]

HI all,

I'm just installing a few switches at a customer.
The deployment is a stack of 2 C9300-24P-M with C9300-NM-8X in the main rack for collapsed core and access at the offices with a few MS130-48P switches scattered around the facility.

When I wanted to aggregate the two fiber uplinks between each MS130 and the C9300 stack I first aggregated them on the access switches before doing it on the C9300's.

When I ultimately enabled them on the C9300 the links went down.

Checking the logs on the dashboard terminal I found these nice entries:
%SPANTREE-6-PORT_STATE: Port Te2/1/1 instance 0 moving from forwarding to disabled
%ETC-5-CANNOT_BUNDLE2: Te2/1/1 is not compatible with Po1 and will be suspended (Broadcast suppression: Level of Te2/1/1 is not configured. Level of Po1 is 1.00%, 1.00%.)
%SPANTREE-6-PORT_STATE: Port Te2/1/2 instance 0 moving from forwarding to disabled
%ETC-5-CANNOT_BUNDLE2: Te2/1/2 is not compatible with Po2 and will be suspended (Broadcast suppression: Level of Te2/1/2 is not configured. Level of Po2 is 1.00%, 1.00%.)
%SPANTREE-6-PORT_STATE: Port Te2/1/3 instance 0 moving from forwarding to disabled
%ETC-5-CANNOT_BUNDLE2: Te2/1/3 is not compatible with Po3 and will be suspended (Broadcast suppression: Level of Te2/1/3 is not configured. Level of Po3 is 1.00%, 1.00%.)

This is strange since the running config on the individual interfaces already had following config

storm-control broadcast level 1.00
storm-control unknown-unicast level 1.00

One of the first things I usually do is limit the BUM traffic by configuring:

So when unbundling those channels the ports came back up as one would expect.

I then disabled storm control first, created the aggregates then and was able to re-enable storm control after the channels were made succesfully.

So for people in similar situations:

- Do the storm control feature in the final steps after the topology is up and running.

[PhilipDAth]

This might not be related, but note this fix in 17.2.1 for MS.

Also note after doing this upgrade you have to DELETE and re-create the existing LAG configurations for the fix to be applied.

[GIdenJoe]

I considered that but in this case it was the first and only link that was being suspended.
And the log message seems to indicate that it is a bug having to do with the creation of the new Po interface.