• Status: Solved
  • Priority: Medium
  • Security: Public
  • Views: 4133
  • Last Modified:

Teaming (LAG) swtich ports across non-stacked switches

Is it possible to LAG or Team switch ports across switches that are not stacked or even stackable?  I say no, that it's not possible even using LACP, but I have a client that believes that it is possible.  

The specific switches in question are the Dell 5424 (a L2 managed switch).  Tech specs link here:
http://www.dell.com/content/products/productdetails.aspx/pwcnt_5424?c=us&cs=04&l=en&s=bsd
0
pwyzorski
Asked:
pwyzorski
  • 7
  • 4
  • 3
6 Solutions
 
kdearingCommented:
Yes, you can.

You can use it to increase bandwidth between 2 devices, or to connect to redundant switches, etc.
0
 
pwyzorskiAuthor Commented:
Yes, I understand the utility of Link Aggregation well enough.

My question was more to the point of how to setup Link Aggregation across non-aggregate switches.  What is the protocol that determines what port on which switch gets used for downlink traffic.  Is it LACP? Static or Dynamic mode?

The protocol to determine which uplink port on aggregate NICs to use for upstream traffic from the host is simple enough to understand.  It's the downlink trip from the network back to the host that confuses me.

I think that the LACP protocol is used to determine a dynamic route and downlink port downstream to the host even across non-aggregate switches. But can LACP be used in static mode to determine a route across non-aggregate switches.
0
 
kdearingCommented:
Enen though 802.3ad is supposed to be a standard, many maunfacturers have implemented their own flavor with 'enhancements' (especially Cisco).

I've set up some Nortel switches and they have a 802.3ad extension called Split Multi Link Trunking.
This allows two aggregation switches to appear as a single device to edge switches that are dual homed to the aggregation switches.

I'm not sure if those specific Dell switches have similar capabilities.
0
Free Tool: IP Lookup

Get more info about an IP address or domain name, such as organization, abuse contacts and geolocation.

One of a set of tools we are providing to everyone as a way of saying thank you for being a part of the community.

 
from_expCommented:
hi!
Teaming can have different types of behavior: aggregated and redundant.
when the first one is generic link aggregation based on 802.3ad, the second use the second card only when the first one fails.

as for switches I haven't seen switches save Nortels where it is possible to use aggregated mode in non-stacked configuration. however it is always possible to use redundant mode with non-stacked switches, where from the switches' perspective you have only two normal ports going to the same server.
0
 
pwyzorskiAuthor Commented:
I read about SMLT in Nortel switches, but that isn't what I have.  I'm more curious about how it's supposed to work in a general sense.
0
 
pwyzorskiAuthor Commented:
"however it is always possible to use redundant mode with non-stacked switches, where from the switches' perspective you have only two normal ports going to the same server."

That statement does not make sense to me.  I thought this to be a routing issue where one switch or the other is chosen as a downlink pathway by a router.  How can a switch know about another switch in a non-aggregated configuration?  Unless you think I'm talking about LAG on a single switch, but I'm not.  The LAG is actually spanned across two non-aggregated switches.
0
 
pwyzorskiAuthor Commented:
Thanks for the responses, but I'm still looking...
0
 
from_expCommented:
no, a bit different
in redundant mode, only server has teaming configuration and only one NIC is active at a time, so you don't need to configured lag on switches.
0
 
pwyzorskiAuthor Commented:
So with the NICs on the host aggregated and no LAG configuration on the switches what we end up with is simple redundancy.  I can get that.

What protocol or protocol aspect defines what switch gets used during a failure event with the primary switch.  Assume simplest scenario with a single router and two switches with multiple hosts connected to both switches.  Is this situation sorted out simply by normal dynamic routing algos?
0
 
from_expCommented:
I would say you even don't need dynamic routing.
static failover only.
so if you have two switches (connected via LAG, without stack) so you can connect each server to both of them having NICs in redundant LAGs and multiple routers connected each to the appropriate switch. two routers then have vrrp (or cisco's hsrp) running to provide gateway redundancy to PCs
0
 
pwyzorskiAuthor Commented:
How about a single router connected to both switches?
0
 
kdearingCommented:
Extreme devices have ESRP which is similar to VRRP/HSRP but also works on their layer 2 switches.
0
 
from_expCommented:
possible, to do that also, but how would you like to arrage failover?
0
 
pwyzorskiAuthor Commented:
some reason client isn't opting for failover at the router level... beats me...

I'm done one this, thanks for the help. I'll award points between you two as such:
'from_exp' - 400pts
'kdearing' - 100pts

thanks
0

Featured Post

Free Tool: Subnet Calculator

The subnet calculator helps you design networks by taking an IP address and network mask and returning information such as network, broadcast address, and host range.

One of a set of tools we're offering as a way of saying thank you for being a part of the community.

  • 7
  • 4
  • 3
Tackle projects and never again get stuck behind a technical roadblock.
Join Now