Archive for the ‘Software-defined Converged Server SAN’ Category

Inside Out Meets Scale Out-Server SAN

Jason Brown

Consultant Product Marketing Manager at Dell EMC

Having recently seen Pixar’s brilliant animation film Inside Out, I couldn’t help but draw parallels between the five fascinating emotions featured in the film and those that affect an enterprise customer’s buying decision, especially when investing in a technology that is as disruptive as software defined. Inside Out takes a refreshing peek inside the human brain where 5 personified emotions – Joy, Sadness, Anger, Fear and Disgust – live in a control room and drive rational responses. It is interesting to see how these emotions can be applied to the software-defined changes taking place today in the storage industry and how EMC® ScaleIO® fares in this evolving market.

Joy & Sadness: Let’s face it! In anJoy enterprise storage setting, Joy equals big bucks – either earned in gross revenue or in cost savings. Just as Joy acts as the prime driver of the personality in the film, all business decisions are driven by a subliminal urge to maximize profit and/or minimize cost. But sometimes the urge for excessive joy overlooks other subtleties in decision making, such as choosing a free, open source solution over a paid but high-performance alternative. An overemphasis on profit or cost reduction has the tendency of skewing away from the strategic trade-off point that gives the optimal returns for your business. As the film evolves into its third act, Joy eventually realizes that Sadness actually performs the critical function of restoring our emotional balance, a key ingredient for decision making. Likewise, as organizations expand and data requirements in terms of capacity and IOPS explode, CTOs and CIOs have to realize that there’s no single magic bullet for their constantly growing storage problems. No one articulates this more effectively than Randy Bias in his insightful take on the charade of ‘unified storage’. Put simply, if you want high performance and scalability, what you need is not a Swiss army knife, but a solution that is optimized and purpose-built for that specific requirement. ScaleIO, which is designed to deliver optimized performance and scalability from commodity servers, plays that role to a hilt.

AngerAnger: The fiery redhead character stays true to its name, acting as a stimulus whenever it sees injustice being meted out. Like Anger, we all want to be dealt with fairness, especially when dealing with huge enterprise purchases. Imagine the frustration of undergoing fork-lift upgrades for end-of-life hardware, with no intention of utilizing the full rack capacity and not having a choice to use it for any other purpose. ScaleIO keeps Mr. Anger calm and composed with its extremely fair and flexible ‘pay-as-you-grow’ scalability and licensing models. In an industry where choice comes at a premium, ScaleIO enables the flexibility of configurations, empowering IT organizations to build the solution that best fits their application and business needs. To top it all, ScaleIO software not only simplifies the deployment of a fully-architected server SAN, it also eliminates the need for unpleasant, rage-inducing activities like capacity planning and data migrations!

Fear: Like an overanxious anti-virus who lives and  breathes FearMurphy’s Law, Fear is on a constant vigil against potential disasters lurking around the corner. In the world of storage, one often finds Fear steering the control panel of our nervous system in case of mission critical workloads susceptible to unscheduled data outages due to outdated IT infrastructure. Enter ScaleIO – designed to survive the unplanned and unpredictable. It maintains uninterrupted I/O operations with its automatic data rebalance / rebuild capabilities across the cluster, and non-disruptive upgrade support. An ESG Lab validation report confirmed that a ScaleIO failure test in the “worst-case” scenario would easily put Mr. Fear permanently and uncharacteristically at ease.

DisgustDisgust: Disgust comes across as someone with really high standards and refuses to lower them. In other words, she can’t stand mediocrity. There are enterprises that are pioneers and market leaders in their own fields and like to maintain that position. They are hard to impress and have high expectations when choosing technology partners. You need to be the best in your product/service category in order to do business with them. When it comes to scale-out, server SAN block storage, ScaleIO is just that. With a tested scalability of over a thousand servers, an eye-popping ten million plus IOPS, and up to 60% better TCO than traditional SAN, ScaleIO is just the kind of technology with whom Disgust won’t mind letting her hair down for a change.

So, with EMC ScaleIO passing the Inside Out test in my books, it’s time for you to download your Free & Frictionless copy of the software and tell us how your control panel reacts!

 

“Live Free or Cry!”…Liberated Data Centers Don’t Cry!

Rodger Burkley

Principal Product Marketing Manager at EMC

Whether you manage a small business; regional office; mid-size enterprise, or full blown enterprise data center with regional/multi-geo sites…data of all types is exploding and costs are increasing steeply. Storage inefficiencies (from innumerable copies, snaps, back-ups and save sets); constant provisioning; data migrations and fork lift upgrades of specialized storage hardware, raw capacity and bandwidth are hemming you in and cramping your style – and data center CAPEX, OPEX and budget. Your hair’s turning white, BP’s up and your spouse or significant other (and the kids and dog/cat) seldom see you.

On top of that, you’re battling vendor lock-in, insidiously escalating shadow IT public cloud usage (and monthly fees) and growing mumbling/grumbling from your less than happy users and SLA clients.  Oh…and, yes, it’s time to go back to the well and ask for even more $$$s as you realize your recently purchased and installed storage array is getting full/utilized at around 50% or so. Time to do a SWAG on how much additional raw storage capacity you’ll be needing over the next couple years and order another sizable array upgrade….and brace yourself for some downtime and tedious data migration(s) and provisions for RAID aggregates, groups, folders and volumes.  Billy Joel’s hit song “Pressure” comes to mind!  Is there a better way to go?  Yes!

Storm Cloud

What can I say, ‘The Cloud’ can be dark and scary as far as your data security is concerned…and it can sneak up on you.

Why not liberate your data center?  Liberate your data center with vendor and hardware neutral commodity based server platforms that monetize your host application servers’ direct attached storage (DAS) for  maximum performance and flexibility.  Also enjoy in the process lower TCO and a new software defined storage architecture that yields true linear capability for compute, storage capacity and throughput.  Liberate your data center with a storage platform that gives you total elasticity to add or remove resources in response to real-world, real-time compute and storage needs with the ability to scale (on the fly, with no down time) to many Petabytes and millions of I/Os per second (IOPS).   So liberate your data center with a software-defined (SDS) server SAN…and EMC’s ScaleIO!

Incidentally, SDS server SANs are optimal for building hyper-scale private, hybrid or public clouds.  Or perfect for building hyper-performance scalable and elastic enterprise storage systems (as in millions of IOPS) for high performance applications (HPAs).  Just start with a minimum of three servers (with DAS of any media type and amount of excess raw storage capacity)…and grow to thousands of nodes with Petabyte scale capacities. Grow at your convenience and need over time – with no disruptions to your operations. Just scale-out with more servers/nodes, repurposed or purpose bought to add to your server SAN..   And scale-up with more DAS devices on those same participating servers. What could be simpler?

The beauty is you won’t miss the extensive upfront planning and justification for additional storage provisioning, forklift upgrades, data migrations, long hours and downtime involved. Just grow in small, ‘baby step’ increments by adding new Flash SSDs and HDDs to existing or new host server local storage….and even create your own storage performance ‘tiers’ (or sub-pools) to match your application or volume’s performance or resource capacity requirements.  Make adjustments to your server SAN cluster using a simple dashboards GUI to monitor system status and enter simple English commands to manage this ‘fluid’ SDS based system.  THIS is the power (and freedom) of software defined platforms. They create an abstraction layer between physical resources and logical management of operations….monitored from a ‘single pane of glass’ application-like software install right on your host server(s).

ScaleIO GUI

ScaleIO’s GUI Dashboard monitors your server SAN’s status, performance and operation in real-time — and tells you when (and how fast) it’s rebuilding data (when servers and/or DAS devices are taken off-line for MX or any reason) or auto balancing across participating host servers.  Simple CLI English language inputs adds/removes host servers and storage devices; designate Protection Domains (for multiple tenants/data partitioning); create Server Fault Sets (to increase high availability and/or scheduled MX) and so on ….. ALL ‘on the fly’ with no service disruption or “outage”….leveraging your DAS excess capacity!

How does this all work?  Simply put, a SDS server SAN system pools together disparate, isolated local storage (DAS) from participating host application servers and forms a single, aggregated virtual SAN that grows bigger and faster as you add more host servers and storage devices to it.  Non-traditional storage….that’s what we’re talking about here.  No specialized storage hardware (host bus adapters, hardware arrays, FC switches or extensive racking and thick expensive interconnect cabling need not apply).  Freedom via commodity hardware-based systems and vendor neutral/agnostic storage solutions….  Live free!

Folks, this is what’s driving hyper-converged storage solutions in the storage industry now…for block (structured data) or file data and object data (unstructured data)  storage. What about the Cloud?  Why rent when you can own….cheaply…using familiar commodity based physical or virtualized servers?  Ever look closely at somebody’s (hopefully not yours) monthly usage utility bill from Amazon Web Services, Microsoft’s Azure or Google’s public cloud?  Be prepared for sticker shock from the ‘we charge you for everything’ usage detail and escalating costs as your “user troops break ranks” and create their own shadow IT operation by subscribing to more and more Cloud services and compute and storage capacity.

And with a robust, resilient and fast performing SDS server SAN you can build your OWN private Cloud on premises….and not have to worry about data security or access when relying on the Public Cloud.

EMC’s ScaleIO is an example of “a robust, resilient and fast performing SDS server SAN.”  OK…so I work at EMC.  Just search for “ScaleIO” on the web and you’ll see what I mean.  YouTube videos walk you through this new technology, architecture and SDS product platform too.  If for no other reason, get up to date on this newest emerging technology that is truly liberating data centers with software based, hyper-converged (all application, CPU, networking and storage functions/resources in one HW platform…or “node”) infrastructure.  Skeptical?  Why not try ScaleIO yourself with EMC’s free and frictionless ScaleIO download. Just like that memorable TV ad a few decades back…”Try it, you’ll like it.”…

Sure, other vendors out there are offering similar SDS product solutions…either via pure software only licensing (usually capacity based) or a pre-packaged/bundled “hyper-converged storage appliance” or both forms. ScaleIO is sold in both forms….software or appliance…and that gives you more flexibility as both formats serve their purposes.  You might also have heard about Ceph and “ICE” (Inktank’s Ceph Enterprises distributed commercially by RedHat).   But when you look at ScaleIO’s inherent high performance, ability to scale from 12TB to 16PB raw capacity, dynamic elasticity and flexibility to add or remove nodes, storage devices and servers without any disruption to operations and other quality of service (QoS) features/functionality ScaleIO brings to your datacenter rack(s)….nothing out there comes close.  Bold statement?  Indeed.  But as they say, “an informed buyer is an educated and confident buyer…” 🙂 (OK…I took some liberties with that saying for sure.)

Does this mean the end to “traditional storage platforms” as we know and (mostly) love them today?  No.  Not hardly.  There will always be a need, use and place for traditional SAN (block), NAS (file) and archival (object/cloud) storage platform products.

The challenge, of course, is in identifying and rationalizing respective use cases, assimilating new technology/architectures and integrating ‘the new and old’ into seamless, efficient deployments that leverage (and monetize) each other’s strengths, specialized features/capabilities and value add that fit YOUR data center and IT needs.   With the storage industry leader like EMC, you’ll be getting all the advice, support, products and solutions (across the entire storage spectrum) to help you take advantage of both newer and traditional storage technologies and value propositions.

In closing, this was definitely an unabashed paid endorsement for EMC’s ScaleIO.  But more importantly, hopefully this post highlights just how much these new SDS server SANs in general are changing the datacenter landscape and freeing IT/MIS CIOs, directors, admins and managers from hardware and vendor lock-in and older, more traditional storage architectures.  It’s happening as you read this….

Want to know more?  Check out @EMCScaleIO on Twitter!   Get on board the emerging storage technology and ScaleIO train.  Live Free…and Don’t Cry!

ScaleIO: Focusing on Total Customer Experience

Jason Sturgeon

ScaleIO Product Manager at EMC

Latest posts by Jason Sturgeon (see all)

Releasing ScaleIO into the Wild with Free and Frictionless

On May 29th 2015, ScaleIO software became available to download by anyone with a couple of clicks to deploy a software defined storage solution on your hardware; with no time, capacity, or feature limits. Sounds simple, and for customers we want it to be simple.  Many things took place behind the scenes to make the release a success that may not be apparent. Initially, we wanted to launch sooner, but based on customer feedback, we knew the installation process could use some improvement. For a customer without prior experience, the installation was not as simple as it needed to be.  We needed to fix that before releasing ScaleIO into the wild. But how did we know what our customers really wanted? Well, we listened. How did we collect the information needed to simplify the installation process?

We started a project to collect Customer Experiencefeedback. We gathered some technical users, and did remote usability testing sessions where we would record them installing ScaleIO. We consolidated feedback about quality of documentation, ease of installation and time to install. Users also had the option to give direct feedback and many did.

We then took all the input to the engineering and documentation teams. What we found was that we needed to change the way installations worked for simple deployments and this led us to create an installation wizard that guides new users through the process. The advanced deployment option is still there for users that prefer that installation process.

ScaleIO Installation

But did it work? Before releasing to the world, we wanted to verify. We ran through our installation testing with a new set of users, using a beta build of ScaleIO version 1.32.0. We also greatly changed how the quick start guide helps a user though the setup. In this second round of testing, we found the feedback was much more consistent and a huge step in the right direction. The time to install went down considerably. They now found the quick start guide cleaner and the steps easier to follow as well.

Interested in becoming more involved?

We’ve got some very cool things we are working on to make the entire experience with ScaleIO even better, so stay tuned. As we continue updating ScaleIO, we want to shape it into something that makes your job even easier. Our best method to do that is to listen to you, so please give us feedback. The first step is to actually take ScaleIO for a spin, if you have not already. Once you have, go to our ScaleIO community site and join the discussions. We are reading and responding to your posts, taking them into account as we evolve ScaleIO with new features and expanded functionality.

Read more about EMC’s Total Customer Experience Global Celebration here.

Why Purpose-Built Storage Still Rules Over “Unified Storage”: How ScaleIO Spanks Ceph on Performance

Collectively it’s clear that we’ve all had it with the cost of storage, particularly the cost to maintain and operate storage systems.  The problem is that data requirements, both in terms of capacity and IOPS are exploding and growing exponentially, while the cost of storage operations and management is growing proportionally to those data needs.  Historically the biggest culprit is “storage sprawl” where we have pairs of arrays throughout the datacenter, each of which has to be managed individually.  Silo after silo, requiring specialized training, its own HA and resiliency, monitoring, and so on.  It’s for this reason that many turned to so-called “unified storage.”  This, unfortunately, is a terrible idea for larger deployments and those running production systems.

Let me explain.

The Storage Unicorn & a Rational Solution
We all want what we can’t have: a single globally distributed, unified storage system, that is infinitely scalable, easy to manage, replicated between datacenters and serves block devices, file systems, and object, all without hiccups.  Bonus points if you throw in tape as well!  This is the Storage Unicorn.  No such beast exists and never will.  Even before the EMC acquisition of Cloudscaling I was talking about these issues in my white paper: The Case for Tiered Storage in Private Clouds.

The nut of that white paper is that tier-1, mission critical storage, is optimized for IOPS, while tier-3 is optimized for long term durability.  Think of flash vs. tape.  These are not the same technologies nor can they serve the same purpose or use cases.

A multi-purpose tool is great in a pinch, but if you need to do real work, you need a purpose-built tool:
Multi Purpose Purpose Built

The issue remains though; how do we reduce storage management costs and manage the scaling of storage in a rational manner?  There is no doubt, for example, that the multi-purpose tool is lower overhead.  It’s simply less things to manage. That is a pro and a con.

The challenge then is to walk away from the Unicorn.  You can’t have a single storage system to solve all of your woes.  However, you probably don’t have to live with tens or hundreds of storage systems either. (more…)

ScaleIO: When More is Better…and Less is More!

Rodger Burkley

Principal Product Marketing Manager at EMC

By now, you’ve no doubt heard how Software Defined Storage (SDS) and hyper-converged server SANs are reshaping the storage industry (and use case) landscape.  Market acceptance continues to grow at an accelerated pace.  There are now many established and new vendors offering hyper-converged storage appliances and SDS systems.   EMC’s own ScaleIO SDS platform continues to score mounting recognition, kudos and accolades with our growing list of customers.Hyperconverged  ScaleIO has even won-over storage techies and IT Management who were initially skeptical that ScaleIO would actually completely deliver on its value proposition and capabilities.

Why all the buzz and hype?  Customers – and the IT datacenter and storage market — get it and are embracing this new, “disruptive” technology by introducing it to their enterprise data centers or using it to create hyper-scalable virtualized infrastructures for cloud applications.  The promise and appeal of installing software on individual commodity host application servers to create a virtual storage pool (i.e., “server SAN”) from each participating server’s excess direct attached storage (DAS) without requiring additional specialized storage/fabric hardware is alluring…and almost too good to be true.  High Availability, commodity based hardware, linear performance and storage capacity scalability, “SAN-less” data storage, lower TCO, elasticity and flexibility are just a few compelling arguments in favor of hyper-converged architectures in general and SDS system in particular. (more…)

Introducing CoprHD. EMC Changes the Game for Software-defined Storage Automation and Management

EMC is certainly no stranger to open source. EMC and Pivotal are both founding members of the CloudFoundry Foundation. And EMC recently announced a $10 million investment and its first CloudFoundry dojo, based in Cambridge, MA, that will attract developers and facilitate the creation of applications on CloudFoundry.  In November, EMC announced the EMC OpenStack Reference Architecture Partner Program and partnerships with Canonical Ubuntu, Mirantis and Red Hat.  EMC also recently launched EMC {code} – the Community Onramp for Developer Enablement, which provides both EMC and community contributions of open source code, drivers, tools, samples, and more. EMC supports and contributes to open source in a number of ways, yet EMC is still considered a proprietary vendor. Well, if none of the above proves EMC’s open source bona fides, perhaps this will: On May 5th, EMC is moving EMC ViPR Controller development into the open source community.

This is big news. For the first time, EMC is taking a commercial product and releasing it to community-driven development.  The open source project, named CoprHD, makes the code for ViPR Controller – all the storage automation and control functionality – available in the open source community. Customers, partners, developers and other storage vendors can download, expand and contribute to CoprHD. EMC will continue to sell EMC ViPR Controller as a commercial offering enhanced with service, support, training, and more to help organizations quickly adopt software-defined storage.

It’s been an amazing journey. Two years ago, EMC announced and subsequently launched EMC ViPR Software Defined Storage. Two years later, The ViPR Controller code, now open source project CoprHD, will be open and available for download on Github.  This signifies a fundamental change to EMC’s development model. All development for ViPR Controller and CoprHD will be done in the open source community, with EMC and others contributing.  CoprHD is licensed under the Mozilla Public License 2.0 (MPL2.0), which encourages community sharing and requires anyone who modifies the source code to share those modifications with the community.  EMC is also establishing free and frictionless access to CoprHD to facilitate community-driven collaboration that will accelerate and expand functionality and support for third party storage.

Why is EMC taking this step? EMC fundamentally believes that software-defined storage is a strategy, not a product. The goal of software-defined storage is to give customers choice of storage services and hardware platforms, make it all simple and less costly to manage, and eliminate proprietary lock-in. Making the ViPR Controller source code available as open source project CoprHD will accelerate development and increase support for non-EMC storage arrays and data protection technologies. It also strengthens CoprHD as a single, vendor-neutral API control point for software-defined storage automation.

This open source model of open, collaborative development is crucial to the future success of software-defined storage and storage automation and management. CoprHD and ViPR Controller will give customers choice, flexibility, and transparency. Purpose-built storage platforms from EMC and others will always remain data center necessities. But customers increasingly value more plug and play architectures – driven by software-defined solutions and standardized infrastructure – and will often sacrifice some level of efficiency to obtain best-of-breed features, more flexibility and lower switching costs. In the modern data center, successful storage vendors will compete on the merits of their solutions and deliver compelling customer experiences. As CoprHD and ViPR Controller extend support to more and more storage platforms, EMC welcomes this new competitive playing field. EMC is ready to lead in this new software-defined world.

Are you a developer that has contributed to a product in the open source community before? Are you planning on contributing to CoprHD?  Are you a storage administrator or architect looking to evaluate and deploy CoprHD? If so, tell us about your experience!  We invite you to join us on this new journey and share your discoveries…let’s see where it takes us.

Categories

Archives

Connect with us on Twitter