2012-05-10 02:32 AM
I am new to NetApp we have a NetApp 2020A so that's 2 1GB NICs per controller, 12 * 300TB disks. This box is going to be used for NFS for VSphere solely, my question is on the best way to setup networking for HA and performance?
At the moment I have 2 VIFS one per contoller setup in multimode. The disk is carved up with 1 controller per 6 disk aggregate. I have 2 NICs per VSpere host. I have 2 * stacked 3Com 5500 L3 switches. To confirm the VIFs are setup as below:
VIF1 10.36.2.100 - e0a & e0b cont 1 multimode
VIF2 10.36.2.101 - e0a & e0b cont 2 multimode
My question is on performance am I better going with ISCSI? And if so what is the best setup with the above hardware if that is the case?
Also on HA I can pop say e0a out of cont 1 and ping is ok and e0a, my question is what happens if a controller goes offline I will lose connectiviy to the aggregrate/volumes managed by it? Is there any way to avoid this I think it cane done with LACP as swtich level but am unsure as how to do this?
Id apprectitta any help you could give me on the above?
Solved! SEE THE SOLUTION
2012-05-10 03:27 AM
My concern is more on the HA elemenent with the FAS 2020 in that from what I can see its active/active but if a contoller fails the VIF IP is not automatically taken over by the second contoller/vif?
Reading more into I dont think LACP will fail over top other controller the difference between it and multimode is that it there is communication between controller port and swictc ion event of failure with LACP?