Aggregation problem

Solved
Guts
New here

Aggregation problem

Hello,

I encounter a link problem between two MS225 stacks with an aggregate of 2 SFP+

Case 1:
1st link:
SFP+ link: vlan 1,10,21,99
2nd link:
SFP+ link: vlan 40-42
= the link works (pinging devices on vlan 1 and 99 between the two stacks OK)

Case 2:
1st link:
SFP+ link, vlan 1,10,21,99
2nd link:
aggregate 2x SFP+, vlan 40-42

= link down, no ping betwen device in vlan 1 & 99

Is there a particular configuration for SFP+ aggregation (RSTP, LACP...)?

Thank you.

vlan.png
1 Accepted Solution
KarstenI
Kind of a big deal
Kind of a big deal

What you did is a misconfiguration of Spanning tree because the Meraki Switches won't do a per-VLAN STP as the "normal" Cisco switches do.

There will be only one active connection (regardless of single link or aggregate) between both switches.

In you first Case the above link was active and your test was done for systems that are active on the first link. But very likely, the systems on vlans 40 to 42 hadn't worked.

On the second case the active link was the aggregate where only VLANs 40-42 were active and every other VLANs were dropped.

What do you have to do:

Option 1: Bundle all three links in one aggregate and allow all needed VLANs. For a better distribution of traffic I would add a forth link, but that is not necessary.

Option 2: If you want to keep this setup with one aggregate and one individual link, you have to allow the same VLANs on both the single Link *and* the aggregate. With this, the aggregate would be used and the single link would be only a backup connection. But IMO, this setup doesn't make any sense and Option 1 is the way to go.

View solution in original post

2 Replies 2
KarstenI
Kind of a big deal
Kind of a big deal

What you did is a misconfiguration of Spanning tree because the Meraki Switches won't do a per-VLAN STP as the "normal" Cisco switches do.

There will be only one active connection (regardless of single link or aggregate) between both switches.

In you first Case the above link was active and your test was done for systems that are active on the first link. But very likely, the systems on vlans 40 to 42 hadn't worked.

On the second case the active link was the aggregate where only VLANs 40-42 were active and every other VLANs were dropped.

What do you have to do:

Option 1: Bundle all three links in one aggregate and allow all needed VLANs. For a better distribution of traffic I would add a forth link, but that is not necessary.

Option 2: If you want to keep this setup with one aggregate and one individual link, you have to allow the same VLANs on both the single Link *and* the aggregate. With this, the aggregate would be used and the single link would be only a backup connection. But IMO, this setup doesn't make any sense and Option 1 is the way to go.

Hello,

Thanks for getting back to me so quickly.

The initial goal was to achieve redundancy for specific server storage vlans.
I have configured the switches as described in option 1 and it works. 
Redundancy will work for all of my vlans and it might not be worse. Good day.
Get notified when there are additional replies to this discussion.
Welcome to the Meraki Community!
To start contributing, simply sign in with your Cisco account. If you don't yet have a Cisco account, you can sign up.
Labels