Are there performance or other benefits to using npiv with rdms vs. assigning rdms "the old fashioned way" directly to the vm? We have several very large luns that are currently presented to physical boxes. We'll be converting said boxes to VMs shortly and I need to decide which avenue to take. Any experienced users out there with npiv?
There can be a great benefit: if you storage of your switch support QoS based on WWN you can give more priotity to I/O operation of you VM.
I think (but never tried) that could be also a secondo benefit: if you realize a MSCS witch NPIV, maybe is possibile to do snapshot on node (and indeed also do VCB)
Andrea
**if you found this or any other answer useful please consider allocating points for helpful or correct answers
No benefits.
--Matt
VCP, vExpert, Unix Geek
just the management perspective
regards
Jose Ruelas
There can be a great benefit: if you storage of your switch support QoS based on WWN you can give more priotity to I/O operation of you VM.
I think (but never tried) that could be also a secondo benefit: if you realize a MSCS witch NPIV, maybe is possibile to do snapshot on node (and indeed also do VCB)
Andrea
**if you found this or any other answer useful please consider allocating points for helpful or correct answers
With performance, no. With regards to storage visibility, yes. You know which vm's have which storage directly from the storage side. It's easier to manage than RDM's. Unless you keep good records on your side, then it's difficult to know which RDM goes with which vm. You can of course run some processes to get that info, but it's not as easy as getting that info from the zoning aliases.
-KjB
VMware vExpert