vSAN1

We have a plan for 3 nodes, but 3 different locations joined by 10Gb. The idea is to have seamless failover between available nodes with HA and VSAN for storage.  Now from what I understand an essentials plus kit will cover this and we need to get vsan licensing on each socket for storage. Am I missing something, like licensing limitations for this?

Help will be greatly appreciated!

I think you are correct on the licensing.  Essentials plus and vSAN licensing per socket should get you the features you mentioned.

I have a question about your design.  Is your plan to have 3 nodes, with 1 node at 3 separate locations?  Thank you, Zach.

Thank you for your response!

Yes, 3 nodes 3 totally separate locations locations, but on the same campus. So we will be joining them with 10Gbe.

Do you have a quick diagram to share?  I'm curious as to how you would handle redundancy between locations.  Thank you, Zach.

Whole campus access network through site B, so if disaster strikes there, all user access down anyhow. Proper server rooms at site A and C, B might fit a 2U. More worried about nodes going down and power outages.

Read up a bit more and now understand we need support for a "stretched cluster". Sounds like one needs the VSAN advanced license. Wonder if this can still work on the "essentials plus kit" and if there are other HA limitations. We basically need to have HA hand over between sites, and this happens every few months.

It seems like it would work, assuming you get the IGMP setup correctly.  This is filed under just because you can doesn't mean you should.  It's outside of best practice and I don't think it would be officially supported.  Thank you, Zach.

Any recommendations on changes?

Not w/o buying more hosts.  Can you get to 6 hosts?  You can then do the 2 Node plus witness implementation at each location; then set each one as a fault domain.

I would have all three nodes in the central location and do my best to have redundant links to the other locations and then make the services (internet, power, cooling, etc...) at the central location as redundant as possible.  Thank you, Zach.