To start off on a tangent, can I say that hasn’t already been said? Alas, the rookie blogger instinct still lives on in me. BUT WAIT. There’s so much to talk about when you dig into what people are talking about. So let us review the reviews that stood out today.
For those still lodged under a rock and whom fortunately came across a whisper of an easily cracked wifi signal, you might have just heard about the announcement of VFcache. In a nutshell, it’s a PCIe card designed for extremely high performance SAN write-through enabled read cache with the ability to double as RW DAS.
That’s a mouthful. But the question is this: Why?
Chuck digs into the only thing that matters at the end of the day – customers wanted to see it:
The best part is that customers don’t have to decide ahead of time which is the “better” approach: server-side or array-side. One environment enables them to use the mix that works for their environment: performance, cost, availability, etc.
Choice is good.
Backed up by the technical enthusiasm that is Chad Sakac, you get the idea – flash is a disruptive force when you look at the performance charts. Mark shines more light by quoting Pat Gelsinger on the release.
But what are non-EMCers saying?
For those still full of curiosity on the specifics, Greg Shultz did a great schizophrenic Q&A with himself on the VFcache specifics. His conclusion?
Thus with VFCache, EMC is using SSD as a cache enabler to discuss the decades old and growing storage IO to capacity performance gap in a force multiplier model that spreads the cost over more TBytes, PBytes or EBytes while increasing the overall benefit, in other words effectiveness and productivity.
On a less enthusiastic tone, Stephen Foskett got into the mix with his stance. He starts and ends pointing to the simplicity of the initial release of VFcache though gets to a point no one would disagree with:
But EMC’s entrance into the host-based flash storage market is a powerful demonstration of the wave of disruption caused by flash-based storage and high-performance computing.
If your mind is like most, you leap to the natural comparison. Luckily The Register ran an interesting article jumping into the lion’s den of WWFioD. Fusion-io’s CEO is said to respond with a simple conclusion:
Cache is a useful function for flash in the server, but cache alone does not address the full potential of flash for application acceleration.
And that’s without a doubt.
Many will dig into the intricacies between Fusion’s DAS solution and VFcache’s split-card mode, SLC vs MLC and the overall commoditization of NAND technologies, but I’ll leave that to those with more time and clearer metrics than I have at the moment. The main point is that everyone agrees with David Flynn’s conclusion or, said otherwise, no one is saying caching alone rids the world of application bottlenecks.
When it comes down to it, your knee-jerk instinct is still the right take away: EMC is now inside your host.
And that, my friends, is big news.