Each host needs access to each controller. (Which looks like your drawing (I think))
example: Host 1 port A-> NetApp-01 port e0c
Host 1 port B -> NetApp-02 port e0c
Host 2 port A -> NetApp-01 port e0d
Host 2 port B -> NetApp-02 port e0d
For Aggrs / vole / LUNs in this config... I’m going to assume you have 24 drives in there, you’d want to creat 2 aggrs, one on each controller. From there, the basic way is just do a volume / lun off each aggr and map to the hosts. Install VSC too, It’ll making mapping datastores easy.
and always test failover on each controller before it goes production.
I would go a step further. e0c/e0d are using the same ASIC. if the ASIC fails, both ports fail. I would use e0c to go to one host on both nodes and e0e to go to the other host on both nodes.
The best solution (as always in a case like this) is to get a small SFP+ based switch and just hook up with Twinax cables.
Be sure to:
use a different subnet on every connection:
Node 1 e0c -> 192.168.10.x
Node 1 e0e -> 192.168.11.x
Node 2 e0c -> 192.168.20.x
Node 2 e0e -> 192.168.21.x
DO NOT use a gateway on the iSCSI networks.
On ESXI, use Port Binding for the iSCSI software ports.
Use Jumbo Frames
Install VSC (in vCenter)
If you are using current ONTAP software (like 9.7), it should allow you to create the aggregates automatically with the GUI (called Provision Storage, if I recall). As @SpindleNinja said, you should end up with two even-sized aggregates, one on each node.
In the NetApp ONTAP world an aggr (aggregate) is a collection of disks that allows data to be written to them. And an aggr is owned by a storage node. like TMAC said, Each port will have its own IP address, on both the storage side and host side. There’s no port trunking/lacp/binding etc with this config.
you want each host to be able to access each storage controller aggr.
Also, here's an example for the last time I did a config like this:
SVM Logical Interface Role Status Network Address Current Port
iSCSI N1_vmhost01a_lif1 data up/up 10.10.10.20 e0c
iSCSI N1_vmhost02a_lif2 data up/up 10.10.11.21 e0d
iSCSI N2_vmhost01b_lif1 data up/up 10.10.10.22 e0c
iSCSI N2_vmhost02b_lif2 data up/up 10.10.11.23 e0d
Host's each have two IP addresses too.
10.10.10.30 and 10.10.11.30
10.10.10.40 and 10.10.11.40