Gain Storage Control with SIOC and Take Performance Control with QoS from SolidFire

Editor's Notes

• #2: SolidFire provides the flexibility and adaptability required to drive large scale VDI environments. SolidFire’s guaranteed performance QoS controlls unpredicable workloads, allowing an administrator to isolate and guarantee performance to each application, eliminating the need to use dedicated storage system for a resource hungry workloads like VDI.
• #4: Use: Table setting Slide
• #5: Add Ultimate Software. Comcast Sungard AS Logo Internap / hosting / datapipe BNPP Lockheed Martin Better mix of ENT customers
• #7: Show the front and back of the nodes Pull element out of this slide Look at moving this up behind the table setting slide Would not break the Element OS off of the HW Use Element as a wrapper / bonding agent that guides the interworkings
• #9: Rapid Response Increase Productivity Managing Real time (add and subtract) Decouple of performance and capacity – managing as a resource pool No end of life, end of sale phenomenon that has plagued this industry for years… don’t have to think about it
• #13: Walk though the bullet points and slide build focusing on failure isolation, self healing and the idea that we don’t depreciate system performance as done within traditional disk based systems leveraging RAID. Be able to discuss and WB Shared Nothing Vs. Shared Disk Architectures, why it is superior and why eBay made the decision to go with SF over other storage systems because of this capability. And why it is a valuable and key requirement for systems at scale Be able to handle the following questions: No RAID!? How do you do that, and why? Why did you implement RAID-less, and why does “shared nothing” matter? NSPOF Non-controller based How do you get faster rebuilds at scale? What does it mean to Isolate failures How is Helix data replication different/better than traditional RAID? How does Helix facilitate non-disruptive outages and upgrades?
• #15: Speaker Notes 3Par has it… Yes they have it Diff = guarantee, 3Par = best effort All bets off if disk fails, node fails etc… To come close you need an 100% SSD platform Why are you not guaranteeing Latency It is about getting what you want out of the system = customer sat Dialing in a satisfaction….
• #18: To deliver predictable and guaranteed storage performance, SolidFire leverages a technology called Performance Virtualization which virtualizes performance resources in very much the same manner that capacity is virtualized today. Patented by SolidFire, this technology allows administrators to manage storage performance separately and independently from storage capacity. With SolidFire IT managers have unprecedented control over their storage system, and are able to deliver predictable storage performance to thousands of applications within a shared infrastructure. Because performance and capacity are managed independently, you can create small volumes with extremely high performance, and very large volumes with very low performance. This allows you to provide predictable storage performance to a very broad number of applications. On the right is a view of how volume level performance is set within the SolidFire storage system Settings can be established for Min, Max, and Burst ensuring that applications get exactly the performance they require (next slide)
• #21: Before / after list Trad mgt tasks
• #22: Floor space = key issue Consolidate 4 racks of Vmax and pull it down to one floor tile Power and cooling Executed across entire Data Set As system grows becomes more efficient, increses overall Decreasing cost / gb Doing more with less Running test dev spin off right next to production Takes up no space Include the visual comparison slide and % to tell the floor tile reduction story Driving cost parity and advantage over disk What the efficiency values actually are… 5 – 10 efficency Proven 7.8x in the field
• #25: (over 2 million API calls per day are being made and growing
• #29: Do not talk about these 5 elements SolidFire was born out of large scale cloud computing at Rackspace Solving a specific problem Informed our target customer Informed our featured set… 1.4PB. 3M IOPS Dave Wright history, born from the cloud Storage array not built by storage people, but built by cloud people. New storage architectures driving changes in storage demands – reacting to the demands driven my changes in compute
• #34: Make Red cover Green Entirely – a little into the blue. Put Arrows from FLASH V2 Not having to choose what applications go on what storage platforms… simplicity Moving away from Silos and the way that it has been done before Fade out the left hand side of the graphic Insert Y axis line Hybrid Array Defense – throwing the sex offender into general population Chris / Fujitzu example