Community Feedback

 View Only
last person joined: yesterday 

How is the community doing? Do you have any questions or feedback related for the Airheads Community team? This is the place to let us know.
Expand all | Collapse all

Aruba 5412 chassis redundancy

This thread has been viewed 16 times
  • 1.  Aruba 5412 chassis redundancy

    Posted Jul 27, 2022 04:01 PM
    What is the best technology to have chassis redundancy using two Aruba 5412 core switches? VSF , Dynamic trunking,...?


  • 2.  RE: Aruba 5412 chassis redundancy

    MVP GURU
    Posted Jul 28, 2022 05:27 AM
    Hi! well...chassis redundancy falls into the case where you have two Management Modules (other than having four Power Supply units) on each Aruba 5412R zl2 Chassis.

    VSF instead creates a Virtual Switch (as seen by the connected peers) logically "aggregating" two Aruba 5400R zl2 chassis (VSF will disable the Standby MM if present, this on each Chassis).

    Currently you would follow two mutually alternative paths: VRRP+DT or, better, VSF. Personally I would go with VSF but it depends on your network requirements and network topology (VSF to be fully effective requires that connected peers are interconnected to both VSF members via LACP links aggregation called Port Trunk in ArubaOS-Switch operating system jargon) otherwise you will not use it fully.


  • 3.  RE: Aruba 5412 chassis redundancy

    Posted Jul 28, 2022 08:16 AM
    Hi , thank you for the answer.
    actually not all peers will be connected via LACP links  , servers and clients would be connected directly to both switches , the majority via LACP links but few clients to be single connected to one of the switches , we do understand that in case one of the switches fail the single connected clients will lose network access but is there any other concerns using VSF while having single connected hosts?


  • 4.  RE: Aruba 5412 chassis redundancy

    MVP GURU
    Posted Jul 28, 2022 10:34 AM
    Hi, well...IMHO the VSF is more aimed at connecting dual-homed Servers or Distribution/Access switches than directly connecting single homed peers (indeed, generally, clients can't be dual homed so easily like Servers and Switches) and so, if you're planning to directly connect edge devices (clients) to one VSF Member only, it is far better, in my opinion, to use a dual homed Access Switch as an additional layer toward the VSF and then connect Clients to that switch (so you avoid to have expensive modules on Aruba 5400R zl2 chassis when you can have more edge facing ports on a cheaper distribution/access switch dual homed to the VSF).

    You asked: "we do understand that in case one of the switches fail the single connected clients will lose network access but is there any other concerns using VSF while having single connected hosts?"

    I believe that there aren't other particular concerns apart the fact that a single homed peer, so a peer connected to only one VSF Member - once that VSF Member goes down or has a fault - that peer will see its single link connectivity totally disrupted and so all the VSF services provided by the VSF and still running on the remaining operating VSF Member (Switching and Routing) can't be served to that particular peer because the peer will become totally isolated...that to explain the importance of dual homed connectivity against a VSF stack IF resiliency and redundancy are among your priorities.