HPE Community read-only access December 15, 2018
This is a maintenance upgrade. You will be able to read articles and posts, but not post or reply.
Hours:
Dec 15, 4:00 am to 10:00 am UTC
Dec 14, 10:00 pm CST to Dec 15, 4:00 am CST
Dec 14, 8:00 pm PST to Dec 15, 2:00 am PST
Web and Unmanaged
cancel
Showing results for 
Search instead for 
Did you mean: 

HP 1950 - STACK with Distributed link aggregation

 
SOLVED
Go to solution
Gustavo_Kaiya
Occasional Contributor

HP 1950 - STACK with Distributed link aggregation

Hi team,

Could you confirm if HPE 1950 switch in stacking mode performs DLA (distributed link aggregation) accross stack members ?

Regards,

Gustavo Kaiya

2 REPLIES
parnassus
Honored Contributor
Solution

Re: HP 1950 - STACK with Distributed link aggregation

Well, hope to not misunderstood your question [***]...but, if for Stack your mean IRF (Intelligent Resilient Framework), and...regarding the HPE OfficeConnect 1950 Switch Series when you speak about deploying Stack (not Management Stacking [*]) you're really speaking about deploying IRF...then up to four HPE 1950 Switches can form a single "Virtual Logical Switch" (the IRF Stack)...once a IRF Stack is formed an incoming Distributed Trunking - as an interconnecting technique among different logical Switches - loses its significance (or, better, becomes not relevant) because no "logical distribution of Trunk links" really takes place at destination...why?

Because IRF Stack Member Switches are not logically separated as normally happens when you work with simply interconnected - trunked - Switches...this is just a benefit of deploying (and working with) an IRF Stack.

Creating a single "Virtual Logical Switch" let the "distributed" part of a Distributed Trunk that terminates on the IRF Stack to become non relevant: the Switch that originates the Distributed Trunk sees just one logical Switch (it doesn't see separated logical Switches as you would normally expect)...this means that each DT Members' Link can freely terminate on every IRF Stack Member [**] without a particular configuration on originating end...this also means that "Distribution" becomes, from the logic point of view of the originating end, irrelevant.

So a Distributed Trunk to an IRF Stack should be seen and configured as a simple Trunk (or, better, as an Aggregated Trunk with LACP, if possible).

Vice-versa - generally - an IRF Stack - seen as a single "Virtual Logical Switch" entity - should still be able to originate one or more Distributed Trunk to different external logically separated Switches (let's say a DT originating from IRF Stack and terminating concurrently to Switch A and to Switch B)...but, probably, this feature is strictly related to the type of Switch used (so a question: does the 1950 support originating a Distributed Trunk when it is deployed as stand-alone unit?)

[*] Note how - on HPE OfficeConnect 1950 Switch Series Datasheet - the acronym IRF is literally absent: documenters called "true-stacking" instead of using the proper IRF acronym...if you then look at the HPE OfficeConnect 1950 Switch Series User Guide (for R3xxx version) you will note that IRF is - finally - called in when the argument is "Stacking" (e.g. page 94 "Stack configuration example" and page 24 "HPE OfficeConnect 1950 stacking (IRF)").

[**] Really you will worry about to physically distribute incoming trunking links within IRF Stack Members to balance incoming connections so if a IRF Stack Member dies the whole incoming trunk survives using remaining incoming links connected to other IRF Stack Members...ideally you would use Port Trunking (with LACP) from the originating Switch.

[***] because, with IRF, there is a difference when one speak about a Trunk which terminates to or originates from the IRF Stack...

Gustavo_Kaiya
Occasional Contributor

Re: HP 1950 - STACK with Distributed link aggregation

Thanks a lot !