Avere is calling out its competitors, alleging that in a bake-off with NetApp, Avere's FXT 2260 appliance offered "sub-90-second boot times" across 145 VMs that existed inside "multiple ESXi servers." Meanwhile, NetApp's comparable hardware had "VM boot times 113 percent longer, with as few as 16 VMs," and only offered a "marginal improvement." Avere doesn't expect you to take its word for it, so the company built out a simple comparison chart to explain all the details.
So what's Avere's secret sauce? Apparently, it's data algorithms. Instead of letting data "thrash," Avere leverages a write-caching technology which "absorbs peak VM Hypervisor writes" by shoving all the important information into "hot data" data stores. When that all-important data is needed, drives don't need to thrash, since the FXT can provide that important data instantaneously.
Or, as Avere explains, FXT allows for …
… aggregated writes [which] minimizes the load placed on the NFS storage and provides relief for boot storms generated by multiple VMS booting in parallel.Avere also promises it can optimize steady-state VMS, where "80 percent of the operations are writes." This allows for overall operational efficiency during extended uptime. And if you're following along at home, you'll have noticed this kind of technology is an ongoing trend in the storage industry -- storage appliances and data solutions are using internalized tiered data structures and sophisticated math to turn high data demand into a no-sweat process. Just take a look at what XIO, Dell and Hitcahi are doing.
It's also important to note that this level of optimization is only a small part of the big story. Be sure to look at how Avnet's focus on the SSD market can speed up everything from storage to virtualization servers. As data center needs inevitably expand, VARs may be called in to do fewer deployments and more optimizations. We'll keep tabs on the optimization trend as cloud and virtualization blossom in 2012.