This message was deleted.

Here are the docs for that : https://docs.harvesterhci.io/v1.4/networking/index

my advice is to use bridge all and use for mgmt... reason I'm saying this is that I have been trying to the opposite and there is always something that prevents you from doing what you would like to do. Like longhorn V2 engine is not using storage network...

@rhythmic-article-81903 Does this mean use 6x10GB in one bridge and assign vlans on top including the one for mgmt ? .

that is how I would do it now. You can even separate that storage network to own vlan if you like

OK. How to restrict storage replication to it's own network ? is it post install config ?

yes. https://docs.harvesterhci.io/v1.4/advanced/storagenetwork