True enough, but if an entire LUN fails you are worried about restoring the LUN, however one very important point:
Say you format a hard drive, you have to wait until the format is finished before you can start putting files on it.
VMFS, each file gets restored in turn, and you don't have to wait for ALL of them to be restored before you can power the first one on.
I brought up all these issues with my team, and they all said the same thing you said, eventually they were quiet, once they figured out I thought this through thouroughly.
Everyone agreed later, that OLD way of doing things is what they based their initial reactions on, smaller LUNS is a thing of the past.
More drives = better performance, that's documented by just about every NAS company.
VM Ware even recommends keeping LUNS small, but I question what they are basing that on... Especially since they support a 2TB LUN, if there was a problem in the design, why would they allow it? I think the engineers are still of the old mindset, that keeping things smaller is better, but I have to disagree, especially when you consider that the backups are easily accomplished, and it's not a single drive you are restoring, they are individual files....
It's all risk, if a LUN goes down, you lose X amount of VM, that's true. But aren't ALL VM's important?
I realize one or two may not impact the company, whereas more would, but Like I said, you restore the LUN, restore the VM's one at a time, and go from there.
for my purposes, it's much easier to manage, than to worry about sheer nubmers of LUNS.. not to mention differences in configuration..etc.. That is a nightmare in and of itself.
