Nick Winkworth

What do Bach, Beethoven and Hitachi have in common?

Blog Post created by Nick Winkworth on Mar 31, 2016

    

... Stood the test of time? Loved by millions? Well, yes, but there’s something else. They are all great composers.

 

If you’ve been shopping for servers recently, there’s a new marketing buzzword you’re bound to have heard: “composable infrastructure”.

 

Everyone’s talking about it, and all the major vendors are falling over themselves to come out with new products that offer this miraculous new capability.

 

But what is “composable infrastructure” anyway? Simply put, it’s the ability to draw from a pool of hardware resources (CPU, memory and IO) to create a physical server that exactly matches an application’s need - defined by software parameters.

Xeon-E5v4.pngOne of the things that makes this concept so compelling right now is the ever increasing density of compute power. Intel's new Xeon E5 2600 v4 (Broadwell) announced today has up to 22 cores on a single chip! That's a lot of computing power! Wouldn't it be great to be able to divide up those cores among several independent computers? It would certainly allow a more efficient use of the hardware.

 

All this hype and excitement has Hitachi engineers scratching their heads. This has been a core capability of Hitachi Blade servers since they were introduced back in 2006, and in fact the idea goes back even further - back to Hitachi’s mainframe heritage actually (not quite as far back as Bach and Beethoven, though!)

 

We call this technology Logical Partitioning (LPAR for short) and it’s a feature that’s been developed over multiple generations of servers - so by now, not only is it highly stable and reliable, but it has evolved some pretty sophisticated features.

 

 

LPAR capability is built right into every Hitachi Blade’s firmware for great performance and secure isolation of partitions (those are our “composed servers”). You can even use it for free as long as you don’t need more than four partitions (you can create as many as 60 per blade, with additional licenses!)

 

So, what can you do that’s so advanced, you ask.

 

“What if I have changing workloads and want to make my server bigger (or smaller) on the fly, without impacting services to my users?” Hitachi’s LPAR can do it! Using “dynamic mode” LPARs can add or release CPU or memory resources as needed.

 

“What if I need a server that’s bigger than a single blade?” Hitachi has the answer here too. Simply add and link CB520X blades to grow from two to eight sockets, and up to 12TB of RAM. That’s a big pool of resources to compose servers from!

 

“Ok, I have to maintain my physical infrastructure, but I don’t want to interrupt the services that are running on my physical servers”. Hitachi LPAR even has that covered. Thanks to years of dedicated Hitachi engineering, LPAR’s capabilities have grown to include features as sophisticated as migrating a running partition from one chassis to another without downtime. Now that's smart!

 

The creation, sizing and management of LPARs can be achieved via a GUI, CLI or RESTful API.

 

While other vendors are telling you about their “cool new capability”, we invite you to take a look at Hitachi’s products and solutions that offer time tested reliability as well as advanced features that they won’t have for years.

 

You can find more information on our website here:
And in this video:

Outcomes