1) Single node device with our concept and architecture is slower then HA. For a simple reason: with MPIO read and write requests are served by multiple nodes @ the same time. So performance increases with the scale-out.
2) We'll represent single + async replication with V8.
kspare wrote:We aren't doing any HA, we just want FAST storage, is there an advantage to using a vdisk over a redirected device?
Well, it pretty simple then - give it as much RAM as possible to run cache, put SW VDs on the fast disks or RAID, and make sure that network part will not be the bottleneck. That should do the work.
Best regards,
Anatoly Vilchinsky
Global Engineering and Support Manager
www.starwind.com
av@starwind.com