CBS350 communication loss after failure of active stacked switch

[P-hazeborg]

Dear,

We have a stack of 3x CBS350 switches, are stacked in native stacking all switch same type: CBS350-24P-4X . The switches are connected with a single stack ring network. We configured fixed unit numbers. The system is working fine.

While testing we switched off the active switch within the stack. We noticed a communication loss between 10 - 60 seconds, before the standby switch took over the active role. We did try it multiple times 

What is the expected communication loss in this setup? Is there any possibility to reduce the communication loss?

[KJK99]

To my knowledge, there isn’t any way to control or even view the innerworkings of the CBS350 stacking. Also, I do not really know why CISCO uses the terms ‘active’ and ‘standby’ there. To me, all switches in a CBS350 stack are active. I have a native stack of 2 CBS350 switches and several other switches and servers connected to the stack. All connections are LACP LAGs and I can see that both links in each LAG are actively used. I’ve done several tests involving restarting, shutting down, pulling out plugs and haven’t seen much of traffic interruption, probably thanks to LACP.

[P-hazeborg]

Kris thanks for you reply.

I dont see an option to select them all as active members also not via a CLI commands. How to select the stacked links to LACP for the stack connections or do you mean the other switches (outside of the CBS350 switches) are connected via LACP?

[KJK99]

I did not mean LACP on the stacking links. It's LACP on the LAGs with other switches/servers.