Transcript Slide 1
Enterprise Storage Reinvented Stanislav Dzúrik FTSS Storage IBM Slovensko [email protected] The IBM XIV Storage Traditional Enterprise Storage Solutions Scale Up Building blocks: Disks/Flash Cache Controllers Interfaces PERFORMANCE Interconnects With this legacy architecture, scalability is achieved by using more powerful (and more expensive) components with higher energy consumption Interface Interface =$+E Cache RELIABILITY SCALABILITY Controllers • • • • • Interface JBOD JBOD IBM XIV Storage Architecture – A Disruptive GRID Design principles: • • • • • Scale Out Massive parallelism Granular distribution Off-the-shelf components Coupled disk, RAM and CPU User simplicity Interface Interface Interface Switching Data Module Data Module Interface Interface Switching Data Module Data Module Data Module Data Module Data Module IBM XIV Enterprise Storage Solution • IBM XIV Storage is based on the following basic principles: • The entire system is one Virtual space • Simple storage provisioning an thin allocation • “Self Healing” - the failure of a component is automatically fixed with no impact on the reliability and performance of the system • “Self Tuning” - the provisioning and management of data should always result in the optimal use of available space • The speed of data access is not dependent on the speed of the drives (no disk hotspots) • Make use of readily available standard components • “GREEN” - Efficient use of resources: Power, Cooling, Space • Best in class TCO - Cost effective IBM XIV Storage Distribution Algorithm • Each volume is spread across all drives • Data is “cut” into 1MB “partitions” and stored on the disks • XIV algorithm automatically distributes partitions across all disks in the XIV disks behave like connected system pseudo-randomly vessels, as the distribution algorithm aims for constant disk equilibrium. Thus, IBM XIV’s Storage overall Interface disk usage could approach 100% utilization when loaded Interface Interface Switching Data Module Data Module Data Module XIV Space Distribution on System Changes • Data distribution only changes when the system changes • Equilibrium is kept when new hardware is added • Equilibrium is kept when old hardware is removed • Equilibrium is kept after a hardware failure Data Module 1 Data Module 2 Data Module 3 Data 4 NodeModule 4 [ hardware upgrade ] XIV Distribution Algorithm on System Changes • Data distribution only changes when the system changes • Equilibrium is kept when new hardware is added The fact that distribution is full and automatic ensures that all spindles Equilibrium is kept after a hardware failure join the effort of data re-distribution [ hardware failure ] after configuration change. • Equilibrium is kept when old hardware is removed • Quick recovery from failure. No disk hot spots - all drives are used equally Data Module 1 Data Module 2 Data Module 3 Data Module 4 Storage in a grid Host Host Host Host Host User Switch (FC/Ethernet) • Efficient and Green by design • Simplified architecture • Each volume a spread on all the drives • Use of large SATA Disks • Integrated software • Thin, smart and simple to manage • Simple migrations • How does it work? Interface Module Interface Module Internal Switch 1 Data module Data module Interface Module Internal Switch 2 Data module • Data is redundantly spare space are spread over all the drives, with parallel access and smart caching to match the performance of high end systems • If a drive fails, the system replicates the lost data across the other drives. System is fully redundant in less than 30 min with minimal performance impact. Making Storage Efficient • Comparing raw capacity to actual application data capacity in traditional storage system shows an amazing gap • Several factors contribute: • Over-provisioning of space • Backups, Clones, BCVs and Snaps • Orphaned space • The XIV Enterprise architecture makes efficient use of the capacity so that you can meet your needs with far less capacity • Enables top reliability and high and consistent performance with energy efficient high density components (e.g. SATA) Over-provisioning • Traditional systems require upfront allocation of predicted space • Volume resizing in the storage system is complex • Volume resizing is virtually impossible for applications • Inevitable application downtime • The result: • Users tend to pad requirements, leading to over-provisioning • Excessive capacity goes unused for months, years or forever • Excessive floor-space, power and cooling is wasted • You’re stuck with old systems • Can’t benefit from expected decline in price and advances in technology XIV Thin Provisioning Concept • Virtualize volume size • Define virtual volumes of maximal capacity for each application • Map them to much less physical storage • Scale physical storage over time without having to change the virtual volume size • The result: • Application volumes never need to be resized • No application downtime • Buy storage over time, just when you actually need it • Don’t pay for over-estimation • Don’t waste floor-space, power, and cooling on unused space • Keep users happy Thin provisioning yields 20-50% saving in infrastructure cost over a period of time Differential Backups and Snaps • Backups: Periodic copies of entire volumes • Maintain regulation compliance • Restore data upon corruption or human errors • Traditional systems use full volume copy for each backup • Each copy require allocation of the entire volume capacity • The solution: Differential backups through XIV Snaps • Save only deltas from the main volume • No upfront allocation - Allocate space over time when new data is written • Actual savings depend on • Number of snapshots • Application profile Using differential backups yields 15-30% saving in infrastructure cost No Orphaned Space • In traditional architectures some capacity is effectively lost over time • The complexity of volume and performance management • The ever changing applications and their storage needs • The result: Idle storage chunks are scattered in the system • Reclaiming them is more expensive than buying a new system • The solution: Let the system automate volume allocation and management • Maintaining perfect equilibrium across the system throughout its lifetime • Automatically handle tasks such as striping, volume resizing, migration etc. • The result: No space is ever lost 10-20% of traditional systems space is orphaned and will never be reclaimed. With the new architecture, no space is ever lost. Stretching a TB to the Max Wasted Space 100% Orphaned Space 80% 60% Full Backups 40% Thick Provisioning 20% Effective Capacity 0% Traditional System XIV System Real-life capacity gain with XIV • Meet the same functional needs with much less net TBs How Capacity Efficiency Translates to Savings • Serve the same applications with fewer TBs • Fewer TBs means less power • Fewer TBs means less floor space SATA disks to save even more power • The power consumption of a system is the sum total of the power used by its components • Since there are so many of them, disks are typically the biggest users • SATA vs. FC disks • SATA drives provide 2-10 times the capacity • Lower spin rate means each disk requires 25-30% less power • The result: A lot less power is used to drive each raw TB (3 to 15 times less) • using SATA drives is not a compromise • The XIV architecture offers primary storage performance for all volumes • Perfectly adapt to any future changes in volumes and capacity Enjoy the power–efficiency of SATA drives with a supporting architecture that handles tier-1 performance and reliability System Power Usage • Power consumption of a system comparable to XIV is 180– 380W per raw TB • Typically using 146GB 15K rpm disks • Power consumption of an XIV rack is 7.7KW • 180TB raw capacity, 79TB net capacity • 42W per raw TB today • Rack power consumption will not change much with 2TB disks • But capacity will double • Consumption per raw TB expected to drop to 21W The new solution uses 4 to 9 times less power for the same (or better) performance and reliability levels System Power Efficiency kW per Raw TB 400 350 300 250 200 150 100 50 0 Traditional System IBM XIV The result is a remarkable reduction in power usage for each raw TB Power per TB is only Half the Story • Using less power to drive a storage system means • Direct power cost savings • Cooling savings, typically adding 30% - 100% of the system power consumption depending on data centre efficiency • Smaller UPS infrastructure • Power is becoming a limited resource • It’s not just a a matter of paying more • Power companies unable to deliver more power to satisfy the expected growth XIV TCO Advantage Management and migration Cost Ease of Management Ease of Management • System administration is virtually effortless • Complex tasks are handled automatically under the hood • Always optimized • Not prone to human errors • Things an administrator no longer need to worry about: • Planning volume layout, • Optimizing for performance • Tiering data to service levels • Ease of provisioning means better service to users • Instant creation of snapshots makes backup procedures simpler Simple management means fewer administrators can give better service to users Migration • Automatic data migration • Migrating thick volumes to thin provisioned volumes • Online data migration from other Storage arrays • New hardware can be added to the system • Better performance, less power, more density • Outdated hardware can be phased out and removed With the new solution, migration is effortless ! Summary • Enterprise Storage Reinvented Attribute Typical System IBM XIV Storage Advantage Disk Drives FC small factor SATA >1TB 4-9 times more storage (with less power and same performance) Provisioning Thick Thin Avoid over-provisioning, grow gradually over time with the best available technology Backups/Snaps Full Differential Snaps Keep backups lean Volume Allocation Manual Smart Orphaned space is never an issue Management Complex Simple Minimize learning curve and ongoing management overhead Ďakujem za pozornosť Stanislav Dzúrik [email protected]