Rarely there has been a single technology that could have such a deep and wide impact on the data center in the same way Flash has. Transforming your data center can totally transform your business, and storage is key to accelerating the whole process. Creating true agility for your workloads can surely boost the agility of your business.
A recent Wikibon research (http://wikibon.org/wiki/v/Evolution_of_All-Flash_Array_Architectures) pointed out that:
- Prices of NAND Flash SSD’s have been dropping sharply compared to Magnetic HDD prices. In 2016, the effective cost/TB of NAND Flash will be for the first time lower than that of Magnetic disks.
- Over the next few years, Flash is projected to gradually much more cost-effective compared to Magnetic disks – up to 7 times!
- By selecting a Gen4 AFA (All Flash Array) such as EMC’s XtremIO (https://www.youtube.com/watch?v=hS84Kc5eN4I,) applications will benefit from modern scale-out architecture, sub-millisecond response times, data reduction rates upwards of 1:3, etc.
Whatever your workloads are, you’ll be able to respond and react swiftly. Be it databases, analytics, business applications, mixed workloads, etc. Apps run faster. Scale better. Cost less. Gain tremendous flexibility. Empower users. Whatever you need to do you can do it better than you ever knew possible with Flash!
Database & Analytics
Database & analytics workloads deliver highest business value. But the fundamental challenges of SLAs, infrastructure complexity, and myriad, cumbersome, and costly copy services across production, dev, test, reporting, warehousing, and more, have put DBAs and Infrastructure admins in constant struggles. This represents a risk to core business.
With Flash, database workloads can benefit from unmatched efficiencies. The proper scale-out AFA database storage platform would consolidate and accelerate everything. This results in supercharged production, dev/test, and BI/analytics workflows with incredible data reduction and copy services. Business gets predictable scale-out performance and low TCO. IT gets more self-service options for copy data management.
Enterprise Business Applications
Business applications from leading providers like SAP, Oracle, and Microsoft power virtually every business. But they have always struggled with performance at peak times, as well as storage costs, and cumbersome, inter-dependent workflows. Other challenges include painful copy services across the application lifecycle, slow analytics, storage sprawl, and more. And storage is usually the bottleneck.
The proper all-flash array reinvents storage infrastructure, improving business application workflows. Here are a few ways such advantages could be obtained:
- Enterprise apps will run with consistent sub-millisecond latency all the time, meeting the most stringent SLAs, 24x7x365.
- Scale-out AFA allows consolidation of all production and non-production copies for all the apps. Everything across the entire business process on one array, with total confidence.
- With the proper AFA, in-memory array copy service significantly increases workflow agility. Real-time analytics, on-demand dev/test, agile dev/ops, etc. become readily and easily feasible.
- Last but not least, in addition to huge CAPEX and OPEX savings, organizations leveraging proper AFA will see additional financial benefits: reduced application licensing costs and lower server and network costs.
Storage for VDI
Some years ago, VDI seemed like the perfect solution to Bring Your Own Device (BYOD) and desktop virtualization for all employees. But it wasn’t! For starters, cheap desktop storage wound up on extremely high-cost enterprise class systems. Then came the I/O storms, VDI storage meltdowns, and wasteful overprovisioning. Meanwhile users hate their application response time. Furthermore, desktop management is a mess, rapid desktop provisioning and updates are virtually impossible, so are SLAs. This all lead to rapidly rising VDI storage and admin costs.
With AFA, VDI is totally different. Whether rolling out BYOD or Desktops-as-a-Service. Flash allows delivering an incredibly cost effective, agile VDI storage platform with data and application services that drive performance at scale, all while saving capacity, thanks to totally unique data reduction.
MS Exchange Workloads
Running Exchange Server can be complex and time consuming. The fact is, planning a static architecture for any application always has serious challenges: complicated acquisitions and divestitures, unexpected growth in mailboxes, new massive message volume to name a few. Meanwhile, Database Availability Group (DAG) architecture is extremely rigid. And most storage solutions, like direct attached storage, actually make adapting to changing needs even more difficult.
AFA unlocks the agility of Exchange Server deployments. For starters, it fundamentally changes the way the application and databases consume capacity, creating unprecedented efficiency, workflow speed, and consistent user performance. AFA allows gaining all the collaboration and messaging advantages that Exchange was designed to deliver through: breakthrough performance, storage efficiency, and simplicity. Plus, entire Exchange, database, and Sharepoint-related workflows can actually be consolidated and accelerated. The result is amazing user and admin experience, new workflow agility, all at lower cost than ever before.
Flash continues to offer new ways to transform DC’s making them more responsive and agile. As we build up Platform 3 (http://www.idc.com/prodserv/3rd-platform/) and beyond, Flash technology will focus on workloads that require performance oriented infrastructure. Examples include:
- in-memory databases (like SAP HANA, GemFire)
- real-time analytics (like Hadoop, risk management, fraud detection, high-frequency apps)
- High Performance Data Analytics (like genomics, facial recognition, and climate analysis).
EMC is investing heavily in that space, starting with the acquisition of DSSD in 2014 (http://www.emc.com/about/news/press/2014/20140505-06.htm.) DSSD will be available towards the end of 2015 and is potentially a critical enabling component for next generation converged infrastructure offerings targeting I/O-intensive in-memory databases and Big Data workloads. In other words, DSSD is going to manage and store Big Data in a faster, more convenient way than ever before.
Supporting the hardware, a few patents have already been filed for different mechanisms of DSSD, such as its unified-address-space, consistent hashing, and self-describing data storage.