This sounds more like a call to Support would be needed, to truly troubleshoot. What are your current speeds observed for the backup? What are the speeds of the FC HBAs you're comparing it to? How are you backing up this data? What tool are you using to back this up? What do your MB/Sec/IOPS figures look like in the Nimble GUI?
Backups are sequential in nature rather than random and the Nimble array is smart enough to bypass the cache and sequentially backup from the NL-SAS media. It's entirely plausible to maximise the 1GbE or 10GbE links on any Nimble array (CS200 or CS400), however the environment and configuration needs to be optimal in order to do so.
Hi Mark, could you supply more details on one of your backup product tests? You mentioned the hosts have two 1G interfaces... Are those shared for all ESX management, VM and iSCSI traffic? How are you measuring sequential throughout? Look forward to hearing more about your setup so we could help you.
If you haven't done so, please give us a call in support (per Nicks suggestion).
Sorry Guys and Gals,
The performance issues were not caused by Nimble. The Arkeia server was not optimally configured and everything was running over the management network. Stopped investigating the performance issues once it was eventually decided that we are getting the CommVault Simpana Suite. I'll post back once all the kit and software is installed.