HPE 6600 design of square topology with DT-lacp with VRRP AND IPV6 ROUTING

masteritlion
masteritlion used Ask the Experts™
on
Hello,

As the title states we need to implement dt-lacp for our vsphere servers, iscsi will be in use to our msa2040 san as well so jumbo frames all around.

So far this is all straight forward.

The kicker here is we also want to implement a switch pair to function as the routing for our core. we run a dual homed network with ipv4 and ipv6 routing. dt-lacp does not work with v6 so how would the links to the route pair support high availability? STP seems to be the only way? is there a more novel approach i'm missing?
Comment
Watch Question

Do more with

Expert Office
EXPERT OFFICE® is a registered trademark of EXPERTS EXCHANGE®
SouljaSr.Net.Eng
Top Expert 2011

Commented:
I'm confused by your question. It's my understanding that dt-lacp is another flavor of Virtual Port Channels like in Nexus correct? That stated, why would you need them to support ipv6 or ip at all since they are layer 2 connections?  Are you trying to do layer 2 high availability or layer 3 or both?

Author

Commented:
I'm trying to do both.


this switch pair is doing ipv4 L3. however dt-lacp restricts L3 ipv6.
Distributed_trunking_restrictions
so the issue is how would you do HA for v6 while still maintaining HA for v4?

the plan is to have l2 and l3 HA.
Andrew Hancock (VMware vExpert / EE Fellow)VMware and Virtualization Consultant
Fellow 2018
Expert of the Year 2017

Commented:
what design do you have in place for your vSphere Servers ?

how many hosts ?
Ensure you’re charging the right price for your IT

Do you wonder if your IT business is truly profitable or if you should raise your prices? Learn how to calculate your overhead burden using our free interactive tool and use it to determine the right price for your IT services. Start calculating Now!

Author

Commented:
Shared storage for vmotion. fault tolerance, LACP on the dswitch for vm's. iscsi for shared storage with mpio via separate links and dswitches for those.

my question mainly pertains to creating a highly available L2/L3 core. i'd prefer if our HA of vsphere is kept separate from this solution. i.e it is already highly available as much as it can be.  the network has many client devices which need to be able to failover in case one switch went down as well. routing is done at the core, not at the aggregate or edge switches.
SouljaSr.Net.Eng
Top Expert 2011

Commented:
At the core you want to run  HSRP or VRRP depending on the vendor switches that you have. For layer two fault tolerance you want to implement port-channels/LAGs between the switches. Depending on the vendor switch you have you can implement mult-ichassis link aggregation or port-channeling. This will work for IPV4 or IPV6.

Author

Commented:
that's the jist of it. it's the vendor nuances that i'm trying to figure out. the last piece of the puzzle is LACP to the servers. we're wanting to implement dt-lacp which would be a multi-chassis lag.
with HPE, dt-lacp can't be enabled while ipv6 L3 routing is enabled on the switches. which is a problem as we need all these pieces to coexist.
we can add an additional pair of switches to accommodate this if needed.
SouljaSr.Net.Eng
Top Expert 2011

Commented:
Then  how about just do multi chassis at the access layer and layer 3 port channels up from there?
SouljaSr.Net.Eng
Top Expert 2011

Commented:
With the multichassis only being at the access layer, your spanning tree domain is greatly reduced. Now for all the L3 port-channels at distribution and core, you can just run OSPF for L3 fault tolerance.

Do more with

Expert Office
Submit tech questions to Ask the Experts™ at any time to receive solutions, advice, and new ideas from leading industry professionals.

Start 7-Day Free Trial