MS-225 Stack failure senario

[ZeeBoussaid]

hello, i was just wondering if anyone might have experienced this scenario:

my setup MS 225-4 stack connected to DIST. MS450 via 2 aggregated uplinks, my question is if switch 2 or 3 went bad and it's completely down, does the stack will break or switch 1, 3 and 4 will maintain connectivity? im asking because if the stake breaks the DIST switch may see the  aggregated uplinks as 2 separated uplinks and may cause a loop. i do however have distribution ports set to root guard and the MS-225 uplinks to Loop guard, but im not if that will prevent it. all these stacks are in production and i dont want to 'test" this scenario at the moment.

Cheers!

[GIdenJoe]

I have a customer with a stack of MS225-48LP that were partly replaced due to the fan issue.

In the two years they run those switches there have been 2 instances that in the night the stack kind of breaks and all links are blocked due to LACP.

Only a hard reboot solved this issue in each case.

[Bruce]

In theory if you have a single switch fail in the stack then the stack should continue to operate as a stack, so long as you have stacked the switches in a complete loop (i.e. Sw1 -> Sw2 -> Sw3 -> Sw4 -> back to Sw1). Which means your aggregated ports should remain aggregated. However, you do need to be aware that there is a period of time where the stack re-converges depending on the failure.

With regard to Loop Guard and Root Guard, Loop Guard should be enabled on the trunk ports, but there should be no reason to enable Root Guard. Root Guard is to prevent a switch receiving a superior BPDU on a port. If the switches are all within your control you should be configuring the switch priorities to ensure that you know where the root is. There Meraki recommendations for Loop Guard, Root Guard, etc. are here if you haven't seen them, https://documentation.meraki.com/Architectures_and_Best_Practices/Cisco_Meraki_Best_Practice_Design/...

[ZeeBoussaid]

Thank you for the answer.. as for the root guard feature..I’m setting it on the distribution switches since they are connected to a Core switch which I need it to be in the lowest priority. Based on your answer my stack then is configured right in case of a failure. Thank you again.