Deploying Virtualised Infrastructures for Improved Efficiency and Reduced Cost

1. Deploying Virtualised Infrastructures for Improved Efficiency and Reduced Cost Adrian Groeneveld Senior Product Marketing Manager

3. Application-Aware StorageKey Differentiation • Differentiated Storage Services aligned to Application Needs • All LUNs should not be created equal • Highest utilisation Rates in the Industry • Buy less storage - use it more efficiently • Optimal Performance, No utilisation Penalty • Remove the capacity versus spindle decision • Provision Storage like Servers • You like VMWare for servers?...you’ll like Pillar for storage • Dynamic storage resource (re)assignment • Easy to change priorities based on business needs

4. REAL Benefits Pillar Application Aware Storage Increase utilisationRates up to 80% Reduce Administrative Time by 50% Lower Cost of Ownership by 50%

5. Why Virtualise?

6. Traditional Infrastructure Growth Server Proliferation Network Infrastructure Users

7. What Does Server Proliferation Mean? • Footprint • Power • Poor Server Utilisation • Administrative Burden • Upgrades • Changes • Processes • Maintenance Cost • Inconsistent Systems

8. Why Virtualise? • Reduce physical footprint • Flexible performance • Application optimisation • Quality of Service within physical resources • High Availability • Simplified DR • Automation • Future proofed environment

9. Deploying Virtualised Storage for Complete Virtualisation

10. A Brief History of Storage

11. Networked Storage Evolution • Small tactical deployment • Technology acceptance • Increased tactical deployment • Fall in cost of deployment • Strategic consolidation projects • Today – multiple islands of SAN and NAS

12. And Then Came ILM • The promise of cost savings • Seamless movement of data between tiers of storage • Freeing up space on primary storage • Storing data on the most appropriate storage devices • It all seems to make sense

13. Today’s Storage Landscape Multiple Technologies Storage Silos Training Costs Service Costs Admin Time Complexity Expensive to Manage Expensive to Own Low Utilisation Power and Space Point Products Performance falls behind capacity Space constraints Hidden costs

14. What if We Could Consolidate Further? • One storage solution to manage • Simpler provisioning • Faster provisioning • Increased utilisation • Lower power consumption • Less cooling • Smaller data centre footprint

15. Storage for VMware

16. Storage Requirements • Server virtualisation only partially solves the problem • Virtual environments require shared storage for maximum benefit • Every virtual machine/application has different needs • This presents a number of challenges • Performance • Scalability • Utilisation • Efficiency • Reliability

17. Storage Challenges in VMware Environments • Performance • Multiple virtual machines with differing SLAs • Inflexible storage means all machines treated the same • Virtual infrastructures grow faster than expected • Simplicity of deployment further accelerates growth • Storage becomes a bottleneck • More traffic through a single port • Scalability • Typically VMware environments grow much faster than storage can be deployed • Typically scalability of underlying storage is limited and challenging • This encourages over-provisioning

18. Storage Challenges in VMware Environments • Efficiency • The managing of the storage is often more complex than managing the virtual infrastructure • Utilisation • VMware recommends provisioning 100% more storage than you need • Virtual machine snapshots • Memory buffers • Future growth • Best practices encourage poor utilisation • Reliability • Storage system single point of failure • High availability at the storage layer can be expensive

19. Solving the Storage Challenge

20. The Pillar Axiom System • One Single Consolidated System • SAN, NAS or Both • Multiple workloads and performance levels • Modular growth • Scale capacity AND performance • High availability throughout • Six click provisioning • Flexible data protection and disaster recovery

21. Enabled by AxiomONE Software Multiple Tiers ONE Multiple Apps ONE SAN/NAS ONE SATA+FC ONE

22. Enabling Application Aware Storage AxiomONE Dynamic Performance Manager

23. Not All Virtual Applications are Made Equal Reference or Transactional Disaster Recovery? How is it accessed? Random or sequential? Priority over other apps? How often is it accessed? Block or file? RTO? RPO? Consistent restart?

24. Quality of Service for Storage • Mixed workloads on a single system • Prioritises I/O between applications across system • Directs performance based on application importance • Increases utilisation across system • Instant performance tuning

25. What Quality of Service Does Minimum % of Queue Allocation High Low High Archive Medium Low Archive Data Layout & Block Prioritization Bands Logical Volumes

26. What’s the Problem? QoS Level 1 Lower performance means missed SLA Higher performance means wasted money QoS Level 2 The Axiom Approach 10 Application A Performance of Provisioned Storage 5 Performance Application B 0

28. Application-Aware Storage Virtualisation Pillar Axiom Storage Services Virtual Server Environment • Appropriate storage SLA mapped against VM • Performance mapped against virtual application • Storage instantly tuned for peaks and troughs in demand Physical Server Service Level A Virtual Machine 1 Service Level B Virtual Machine 2 Service Level C Virtual Machine 3 Network(Priority) Memory(Cache) CPU(Priority) Disk(Layout) CPU(Priority) Memory(Cache)

29. Quality of Service for Application-Aware Storage

30. Scaling Your Virtual Infrastructure Scalable Capacity and Performance in a Single Virtual Storage Grid Storage CPUs IOPS TB IOPS TB IOPS TB IOPS TB x 64 Bandwidth Capacity IOPs

31. Application Aware Storage for VMware Environments

32. Storage in a Virtualized Environment • Typical approach – one central array treating all virtual machines as equal • Better than direct attached storage but creates problems

33. The Issue of Capacity vs. Spindles • 300GB Drives set as Raid 5 (4+1) provides a 1.2TB LUN • Multiple VM’s are usually supported by this “stacked array” • Result: 100% random I/O, overloading cache The Solution – Application Aware Storage • Each LUN is striped across four RAID 5 disk groups • Spindle count increased to a least 24 disks • Performance of individual LUN optimized for each VM

34. Pillar in a Virtualised Environment Capital Savings • Optimization Removes the Capacity “white-space” • Capacity lost to underutilised Server • Capacity lost to various RAID configurations • Lower Initial Acquisition Cost due to higher utilisation • $/GB Value assigned for each VM • $/IOP Value assigned for each VM • Thin Provisioning reduces initial capacity need

36. Achieving Higher Utilisation • Consolidation • Eliminate multiple physical tiers • Eliminate separate archive and content based storage • Intelligent data layout • More intelligent use of spindles and available capacity • Distributed RAID to deliver better performance from existing capacity • Capacity on demand growth • Intelligent thin provisioning across all data tiers or priorities

37. Utilisation Without Compromise “We are consistently delivering over 62% utilisation of available capacity – and that’s written data”

38. Summary • Virtualisation can make very significant efficiency improvements • Space, power, cooling, utilisation, maintenance, cost • Virtualisation should be looked at from both a server and storage point of view • One has limited effectiveness without the other • Don’t forget to consider scalability • Virtual environments do scale quickly • Quality of Service at the server level needs to be matched by storage level quality of service to be effective • An effective virtual infrastructure will • Drive down costs • Increase administrative productivity • Improve efficiency • Just make life easier

39. Questions and Answers Thank You