Hi,
I am running ESXi 5.5 on 2x R910s connected redundantly to an MD3000i+MD1000+MD1000. Things are running well with one exception:
I am have about 25-30 VMs powered on at any given time. Just last week, all of my VMs took a huge performance hit and the CPU meter on the cluster went from 170GHz free to 200GHz over provisioned. RAM and storage stayed the same. This fluctuated a lot for days. We are in the process of migrating Windows XP VMs to Windows 7, so I deleted 10 VMs and changed the logging at all levels (hosts, cluster, etc..) to NORMAL (was at verbose). I then submitted a ticket to VMware. They got back to me asking if the problem was reproducible. It was not. We then went down the list of hardware I was using and this is how we discovered that the MD3000i was not compatible. He would troubleshoot no further and would only suggest to update all drivers and firmware on all hardware since I was not using compatible hardware (understand, but to completely drop everything due to this is not what I wanted to hear.) It seemed that the 800MB of logs was not even looked at to determine the problem, anyway, back to the point:
It has been working flawlessly since I deleted VMs and changed the logging level. I am going to continue to use the MD3000i for another 15 months until it is in the budget to upgrade. I hope this helps.