Scale IO Software Defined Block Storage

Editor's Notes

• #3: So as you saw, the 3rd Platform (Big Data, Mobile, Cloud and Social) has begun shaping the way organizations do business and as it evolves to include next generation, modern applications, the need for more agile, scalable, flexible and cost effective storage options have increased the need for storage to evolve as well. Technology has been changing and evolving for years with the advent of virtualization and software-defined. When it comes to software-defined, computing infrastructure/server virtualization has been in place for years, networking is currently underway and … now it’s time for storage to take its place. As a result we have seen a rise in the adoption of the Server SAN and software-defined storage. IT organizations are moving away from what we know as traditional storage and they are moving towards software-defined storage and server SAN. The numbers speak for themselves Server SAN is projected to have a 23% CAGR from 2012 to 2026, with faster growth from 2014 to 2020 of 38% CAGR Traditional enterprise storage market is projected to decline by 16% CAGR  http://wikibon.com/server-san-enabling-business-transformation/
• #4: IT organizations are challenged with finding more cost efficient ways to manage the increase in demand and data, while still increasing revenue, reducing risk and delivering services to the market faster. This requires more agility, operational efficiency, and faster decision making. Then there is the demand for cloud-like elasticity and scale from customers and Lines of Business owners who “need” things now. Delivering services in weeks or months is no longer acceptable. People want services delivered in hours or even minutes. <CLICK> These challenges are only compounded when you look at the rapid changes in the market. Digital Data is growing by the tens of zettabytes, which will put a huge strain on IT professionals. In addition, the adoption of modern applications, which will help to further drive your business and maintain a competitive edge, is quickly outpacing IT staff. And finally, you are faced with the question as to whether or not to build a new infrastructure from the ground up to help manage costs, vs. the need to get your services to market faster by simply buying a fully architected system. In order to meet these challenges and continue to grow your business, you will need to begin to transform and modernize your data center. Key points: IT organizations are under unbelievable cost pressure and pressure to deliver services and applications faster. IT organizations requires more agility, operational efficiency, and faster decision making IT leaders are realizing that they must undergo significant transformation across the IT landscape __________________________________________________________________________________________________ Sources: Data is growing exponentially and its structure is changing. There is a huge change in the way applications are developed & written. IDC has predicted : by 2020, the digital universe of data will grow to 44 zettabytes or 44trillion gigabytes. To put this in perspective, if the digital universe were represented by the memory in a stack of tables, there would be 6.6 stacks from the earth to the moon by 2020. (in 2013, we were about 2/3 of the way there) (http://www.emc.com/leadership/digital-universe/2014iview/executive-summary.htm) And by 2018 over 1PB of data will be managed by each IT Pro Gartner reports: Modern Apps are Up 700%, Outpacing IT by 5X. (Sources: IDC, Gartner, AWS Workload Estimates – SDS Playbook) As a result traditional storage is quickly becoming unsustainable. Many businesses and IT organizations are experiencing challenges brought about by the need to mange rapid data growth and while supporting both 2nd and 3rd Platform workloads. Relatable Stats:Forbes (Sept 2015): http://www.forbes.com/sites/bernardmarr/2015/09/30/big-data-20-mind-boggling-facts-everyone-must-read/#6195a4e6c1d3 The data volumes are exploding, more data has been created in the past two years than in the entire previous history of the human race. Data is growing faster than ever before and by the year 2020, about 1.7 megabytes of new information will be created every second for every human being on the planet. By then, our accumulated digital universe of data will grow from 4.4 zettabytes today to around 44 zettabytes, or 44trillion gigabytes. In Aug 2015, over 1 billion people used Facebook in a single day. Facebook users send on average 31.25 million messages and view 2.77 million videos every minute. Within five years there will be over 50 billion smart connected devices in the world, all developed to collect, analyze and share data. By 2020, at least a third of all data will pass through the cloud (a network of servers connected over the Internet). Distributed computing (performing computing tasks using a network of computers in the cloud) is very real. Google uses it every day to involve about 1,000 computers in answering a single search query, which takes no more than 0.2 seconds to complete. For a typical Fortune 1000 company, just a 10% increase in data accessibility will result in more than $65 million additional net income.
• #5: So, what are some of the IT challenges that cause organizations to hit the technology and cost wall? (Note to speaker – here you should talk to the bullets and ask if the customer is currently experiencing any of these) The questions you need to ask yourself, in the face of this new era of innovation and application transformation, are: Can you achieve the profitability you are expected to achieve using the infrastructure you have today? Are your current and projected CAPEX / OPEX too high? Can your infrastructure handle the expected growth without fear of breaking? Can your current cloud grow to handle growth in data and customer needs?
• #6: So, how is technology changing and how can this help you address these challenges? Before we look at what can do to transform storage, let’s look at what we are doing today and why it is not meeting our current and future needs. Prior to the 1990’s all of our mission critical apps and data were running on mainframes but it was expensive, so people began to look for a better way. This brought about the rise of open systems with HP-UIX, Solaris, SGI and all of the UNIX systems. So we attached direct attached storage (DAS) to all of these systems and used a protocol called SCSI. While this helped to lower price, organizations were unable to protect, manage and share their information. Then we quickly learned that not only could we easily share, manage and protect the data but disk drives were expensive and the network was not performant enough. (Islands of DAS) So, the world decided, DAS was not the answer. This lead to the introduction of the proprietary network call Fibre Channel switched fabric. These systems have been valuable to organizations for years and have delivered upon their original goals, however, even these have limitations. Being proprietary built means that your business and applications depend upon dedicated storage arrays. That equals dedicated workloads and a complex network of Fiber Channel and HBAs. Dedicated SANs and dedicated workloads cannot always provide the scale, elasticity and flexibility needed to meet the demands of applications and the business as a whole.
• #7: Transformation is becoming possible with the changes in resource availability. Years ago, the high costs of CPU and memory, along with the limitations of networks drove the need for powerful storage arrays and Fibre Channel networks. However, networks are changing. There is more bandwidth available today and it is becoming more and more affordable and performant than ever before. CPU has become less expensive. As a result how we manage storage is changing. Software-defined storage, hyper-converged, is one such way. Why? SDS built as hyper-converged is the ability to serve up compute and storage within the same x86 server hardware. It also includes Ethernet/Infiniband networking. With SDS you eliminate tech refreshes, data migrations and expensive FC Infrastructure. The first time you go to software-defined is the last time your organization will have to perform a data migration. There is a reason why Facebook, Google, and other major Web organizations have adopted SDS. It works! It delivers great agility for you and your customers – delivering fast, agile provisioning to meet go-to-market pressures. Next-generation apps (Facebook, Twitter, Uber) are using SDS, and they’ve achieved premium market share in their industries. When you go hyper-converged, you also achieve greater cost savings & efficiency, public cloud agility and the ability to scale both storage and compute linearly. You can also grow your storage and compute infrastructure as need arises, rather than trying to anticipate data growth needs. If you want 300TB today, easy just add the server nodes to the environment. 3 months from now you need to roll out a new service and require, 40TB – just add the nodes again. Demand was not as great as you anticipated? Move some of the nodes over to another service that is growing quickly… Add what you need at that time with a solution that is purpose built to provide that level of flexibility and agility. So as you can see there are many advantages to hyper-converged with SDS but it comes with operational changes. Some organizations are not quite ready with the processes and people changes needed to move to hyper-converged. Does this mean you cannot move forward just yet? No, you can still begin to make changes today and achieve the benefits of SDS.
• #8: Modernizing the data center starts with removing your dependency on traditional SAN hardware and instead, transforming your infrastructure with Software-defined Storage. By utilizing purpose built storage software, combined with industry leading x86 hardware, you can design an infrastructure that delivers the agility and flexibility needed to respond and scale as business demands dictate. This is why ScaleIO is the Modern Data Center. ScaleIO is designed to support traditional application workloads and next gen applications. It delivers the capabilities needed to transform. All-Flash – ScaleIO was built to bring All-Flash to the data center in a SDS architecture. It helps organization achieve Petabytes of All-Flash capacity with a highly disruptive TCO Scale-out – engineered to deliver at scale, ScaleIO is the most scalable SD Block storage on the market today. As a result of this architecture data migrations, islands of SANs and forklift upgrades become a thing of the past. Software-Defined – by its very design, ScaleIO embodies the flexibility customers are looking for from an SDS solution. From being OS/Hardware agnostic to empowering IT organizations to deploy as a storage-only or hyper-converged infrastructure – the possibilities are endless. Trusted – With ScaleIO customers are able to utilize IPv6 addresses, integrate ScaleIO with their role-based user management platforms as well as provide more granular access for those users. Additionally, customers gain confidence knowing that the system has internal logic to validate and ensure the security of ScaleIO components.
• #9: ScaleIO is software defined storage designed with enterprise-grade resiliency to deliver, scale-out, block storage using industry standard x86 hardware. ScaleIO provides the scalability, elasticity, flexibility and performance needed to meet the demands for whatever comes next. It is also a key building block in next generation hyper-converged systems.
• #10: ScaleIO is designed to massively scale from three to thousands of nodes. Unlike most traditional storage systems, as the number of storage devices grows, so do throughput and IOPS. The scalability of performance is linear with regard to the growth of the deployment. Whenever the need arises, additional storage and compute resources (i.e., additional servers and/or drives) can be added modularly so that resources can grow individually or together to maintain balance. Storage growth is therefore always automatically aligned with application needs.
• #11: Every server in the ScaleIO cluster is used in the processing of I/O operations, making all I/O and throughput accessible to any application within the cluster. Such massive I/O parallelism eliminates bottlenecks. Throughput and IOPS scale in direct proportion to the number of servers and local storage devices added to the system, improving cost/performance rates with growth. Performance optimization is automatic; whenever rebuilds and rebalances are needed, they occur in the background with minimal or no impact to applications and users.
• #12: The ScaleIO system reconfigures itself as the underlying resources change; data is rearranged and spread evenly on the servers to optimize performance and enhance resilience. All of this happens automatically without operator intervention and therefore eliminates the need for costly and disruptive data migrations.
• #13: When it comes to flexibility, ScaleIO delivers. It offers you the ability to redesign your storage environment using a storage-only, hyper-converged or mixed approach. If your immediate goal is to reduce costs and maximize scale and performance, without impacting your current staff and operations, you can begin with a storage-only software defined server SAN model. Leave your applications running the way they are today, on independent servers, and simply transform the way you handle your block storage. You can also choose to run your applications on the same servers as your storage, bringing compute and storage together within the architecture. You have the flexibility to scale your storage and compute separately to meet workload and application demands. These hyper-converged deployments offer the best TCO savings, while allowing you to modernize your data center with greater efficiency.
• #14: ScaleIO is infrastructure agnostic making it a true software-defined storage product. It can be used with mixed server brands, operating systems (physical and virtual), and storage media types (HDDs, SSDs, and PCIe flash cards). In addition, customers can also use OpenStack industry standard hardware for storage and compute nodes.
• #15: ScaleIO offers a set of features that gives you complete control over performance, capacity and data location. For both private cloud data centers and service providers, these features enhance system control and manageability—ensuring that quality of service (QoS) is met.
• #16: Trying to predict the growth of your company is hard. If you under predict your capacity you need to purchase more, and if you over predict you will be paying for underutilized storage. With ScaleIO, you “pay as you grow”, and therefore you do not need to pre-pay for infrastructure you may not need. It creates a server based SAN, so there are no dedicated storage components like fabric and HBA’s required. If you chose to deploy in a hyper-converged infrastructure, this reduces costs of power, cooling, hardware maintenance and floor space. In a recent IDC study organizations reported that they are deriving significant business value from their investment in software- defined storage with EMC ScaleIO. IDC determined that ScaleIO translates into significant cost savings and agility for our customers and the numbers here tell the story.
• #17: ScaleIO is software that creates a server-based SAN from local server storage to deliver flexible and scalable performance and capacity on demand. It was designed with enterprise-grade resilience as a must-have to deliver the scalability, elasticity, flexibility and performance needed to meet your business needs, today, and your long term goals for the future. Let’s look at the many benefits it offers.
• #18: Lets take a look at how ScaleIO works, visually. Here we have the servers, network, and storage as separate entities, and then we’re showing it all combined into one single server.
• #19: Each of these boxes represents a server with direct attached storage. Each one has its available IOPS and TBs of storage and they all operate separately. Now if I add the ScaleIO client to the server, you can see a virtual SAN is created. Now these 10 separate servers have combined to become deliver the highest amount of IOPS and TBs of storage across the cluster. And as you add servers, up to thousands, the number of IOPS and storage will continue to grow.
• #20: In a Storage-only Architecture: The SDC exposes ScaleIO shared block volumes to the application. Access to OS partition may still be done “regularly” ScaleIO data client (SDC) is a block device driver The SDS owns local storage that contributes to the ScaleIO storage pool ScaleIO data server (SDS) is a daemon / service In a Hyper-Converged Architecture: The SDC & SDS reside within the same host An SDC and an SDS can live together. SDC serves the I/O requests of the resident host applications. SDS serves the I/O requests of various SDCs.
• #21: With ScaleIO, storage and compute resources can be increased or decreased whenever the need arises. The system automatically rebalances data “on the fly” with no downtime. Additions and removals can be done in small or large increments. No capacity planning or complex reconfiguration due to interoperability constraints is required, which reduces complexity and cost.
• #22: The ScaleIO system reconfigures itself as the underlying resources change; data is rearranged and spread evenly on the servers to optimize performance and enhance resilience. All of this happens automatically without operator intervention and therefore eliminates the need for costly and disruptive data migrations.
• #23: This slide shows the power of the distributed architecture of ScaleIO. Every SDC knows how to direct an I/O operation to the destination SDS. There is no flooding or broadcasting. This is extremely efficient parallelism that eliminates single points of failure. Since there is no central point of routing, all of this happens in a distributed manner. Each SDC does its own routing, independent from any other SDC. The SDC has all the intelligence needed to route every request, preventing unnecessary network traffic and redundant SDS resource usage. This is, in effect, a multi-controller architecture that is highly optimized and massively parallel. It allows performance to scale linearly as the number of nodes increases. ScaleIO is capable of handling asymmetric clusters with different capacities and media types. A fully converged configuration will have even higher parallelism and load distribution between the nodes. Any combination of the fully converged and two-layer configuration options is valid and operational. Note to presenter: Metadata is not in the data path – removes the scaling limitation that exists within other architectures.
• #24: EMC ScaleIO has delivered upon the promise of software-defined, scale out, block storage on standard server hardware with extreme performance, massive scale and superior elasticity. Now you can choose how you wish to consume ScaleIO. There are 3 Consumption Models available. ScaleIO Software: Ultra Scale-out SDS Software Only Complete flexibility End user supplies server End user supplies switch End user supplies rack ScaleIO Ready Node: Scale-out Block Storage Node Dell PowerEdge Servers tuned, optimized and validated for ScaleIO Hyper-converged or Storage only All-Flash configurations VCE VxRack FLEX: Turnkey Fully Engineered Solution Fully productized platform VCE factory integrated & logically configured VCE Support & lifecycle assurance
• #28: 1.  Visit EMC.com/ScaleIO where you’ll have access to whitepapers, videos, demo’s etc.  (www.emc.com/scaleIO) 2.  Join the ScaleIO product Community https://community.emc.com/community/products/scaleio 3.  Download the Free & Frictionless version of ScaleIO http://emc.com/products-solutions/trial-software-download/scaleio.htm 4.  Reach out to local sales team.