Having been one of the very first Scale Computing partners worldwide to beta test Scale Computing’s Hyperconverged platform back in 2012, and one of the first (if not the first) UK service provider to deploy HC3 clusters in a commercial private cloud capacity for our APP:Host solution, we at ACC are really excited to finally be able to publically talk about Scales two new SSD enabled node offerings, the 2150 and 4150 nodes. These two new node types have been released in conjunction with Scale HyperCore (Scales Node O/S) version 6.6 which is the first outing of the hotly anticipated H.E.A.T. (HyperCore Advanced Automated Tiering) system. This joint release truly transforms Scales HC3 solution from a rock solid Hyperconverged platform, to one that is both rock solid and now blazing fast too.
The ‘rock solid’ aspect here is well worth dwelling on. Speaking as both a Scale Computing Platinum partner, and an operator of many Scale HC3 clusters (for some seriously large organisations), when Scale Computing added SSD tiering (that is full SSD tiering, not simply SSD caching) into their HyperCore platform, they did so with their typically intense focus on stability & reliability. Just like with the roll out of all the major platform upgrades we have been a part of before, Scale resisted the temptation to rush H.E.A.T. to market, instead electing to get it right first time. To have sat through H.E.A.T. development conference calls where the main concern is not how to get the product to work, but how to get the product to break – paints a picture perfect illustration of Scales attitude to product design.
So – with H.E.A.T., the Scale cluster operator (and yes, it could quite easily be a single person) when managing a cluster built using the latest SSD equipped nodes, he or she is able to easily assign % block heat map cover to SSDs for any virtual disk. This could be set to 0 (nothing on SSD) or ‘11’ (in a cute homage to the cult movie This Is Spinal Tap) where the entire disk is SSD based. See the short video below – as is standard with Scale HC3, H.E.A.T. is an extremely easy to use & simply configured tiering solution – there is not much more to actually write about, which neatly I think highlights what for us is the key USP of HC3s entire value proposition.
Are we excited about H.E.A.T., and full SSD tiering? Yes!
Are we excited about the performance these features deliver to the platform? Yes!
Is H.E.A.T. the main reason why you need a Scale HC3 cluster running your VMs. NO!
No – actually, it’s not, and let me explain why:
ACC have been running multiple production HC3 clusters as part of our APP:Host solution since HC3 was launched way back in 2012 (i.e. for a LONG time), and as mentioned above, were the first UK hosting provider to do this, so impressed were we with both the product, and the company behind it. Since switching hosting operations over to HC3 clusters, our daily workload for operating & managing the hosting infrastructure - simply vaporised overnight. Gone. No more SANs problems, no more server issues, no more VM software and their clunky management systems. So yes, HC3 is the ultimate in convergence of hardware, software and reduction in running costs – and yes again, H.E.A.T. now delivers a tremendous performance boost to the platform - but for us, it’s the emergence of something even more valuable – time, and it’s this, not H.E.A.T., which should be your key driver to purchase. The HC3 platform enables ACC to devote pretty much 100% of our time on what is most important to us, what’s running inside servers rather than the infrastructure running under them. On the server is where we want to be spending our precious time, because this is the place where our customers get to appreciate our skill at first hand, and therefore what makes us a far more attractive application hosting provider than, say, having your server stuck up in the cloud, along with a couple of hundred thousand others, in some anonymous datacentre who knows where.
So the point here is that there are clearly two routes to simplify your server hosting IT operations:
1) You can outsource everything to a cloud provider, hope it all works, hope the provider understands your business, and hope they care about your data as much as you do.
2) Or you can follow our lead, retain ownership your most valuable asset, your data, and put it on Scale HC3 clusters instead.
H.E.A.T. is a neat acronym that Scale invented for their new SSD tiering system, but many years ago, I invented my own acronym for the word HOPE, which is just as true today.
Those who depend on H.O.P.E. to run their organisations’ tech ops will inevitably be….
P45 (or Pink slip if you are from the US!)