Innovation storage technology for Business Continuity Andrey Pilipenko Senior Solutions Consultant November 2014 Customer Challenges and Market Dynamics Customer Challenges Site Failure ‒ Catastrophic events like an airplane taking down a datacenter & recovery issues Business Challenge ‒ Non-Stop Operation ‒ Reaching zero RPO & RTO Limited Failure Scenarios Gaining Connectivity from Surviving Cluster BUSINESS CONTINUITY Market Dynamics Game changer for customers needing active-active compute and storage configurations Global-active device will allow customers to move to an activeactive fully fault tolerant storage configuration ‒ All that is required is global-active device, multipathing software, and VSP G1000 systems, excluding the quorum. ‒ Enterprise-class customers seeking the strictest IT availability requirements will gain from global-active device when failure is not an option New HDS Storage Main directions NONDISRUPTIVE MIGRATION Active NEW Continuous Seamless identity absorption and transfer 3X Performance UNIFIED MANAGEMENT Active - Active Virtual Storage Workload Mobility Machine Agility Seamless Availability OLD OLD INTEGRATED ACTIVE-ACTIVE Site A Site B Extensible Self-Managing Policy-Driven Streamlined App Level SLA and Business Customization Zero Downtime Flexible Mobility Unified File and Block Continuous Cloud Infrastructure Foundation: Powered By Hitachi Storage Virtualization Operating System Virtual Storage Platform G1000 Redefining Mission-Critical Storage Virtualization Command Suite V8 Hardware Product Packaging Unified Block and File Configurations File Modules Add up to 8 nodes to address customer’s file sharing needs • Up to 2 controller chassis Block-Only Configurations Entry, Small, Medium, Large, Extra-Large Models Flash-Only Configuration • Up to 6 racks • Same upgrade flexibility options as with VSP Storage Virtualization Operating System Mainframe-Only Package VSP G1000: Block Storage Specifications VSP G1000 VSP G1000 Single Chassis Dual Chassis Maximum flash devices 96 FMD 192 SSD 192 FMD 384 SSD 192 FMD 256 SSD Maximum internal drives 1152 2.5-inch disks 576 3.5-inch disks 2304 2.5-inch disks (2.7PB max) 1152 3.5-inch disks (4.5PB max) 2048 2.5-inch disks (2.4PB max) 1280 3.5-inch disks (5.0PB max) Data Path 384GB/second 768GB/second 128GB/second Control Path 64GB/second 128GB/second 64GB/second Maximum Virtual Storage Director pairs and total # of cores 4 VSD pairs Total of 64 cores 8 VSD pairs Total of 128 cores 4 VSD pairs Total of 32 cores Maximum Host ports (if no BEDs) 64 (96) Fibre Channel 64 (80) IBM® FICON® 80 FCoE* 128 (192) Fibre Channel 128 (176) FICON 176 FCoE* 192 Fibre Channel 176 FICON 88 FCoE Maximum Cache Memory 1TB 2TB 1TB Power comparison: Dual controller, (4) VSD pairs, 1TB cache, (2) BEDs, (2048) 900GB drives, 160 Fibre Channel ports - 30.1 KVA 34.0 KVA Max Local copy pairs 32K 32K 16K Max Remote copy pairs 64K 64K 32K LUNs/LDEVs 64K 64K 64K Maximum Bandwidth VSP Simplified Data Center Planning AGILE Deploy to Fit Data Center Requirements Increase floor space efficiency Eliminate data center hotspots Flexibly scale performance and capacity Traditional System Layout Today: VSP G1000 Flexible Deployment Separate Storage Controllers 5M, 30M, 100M Options Coming Soon: Ultimate Deployment Flexibility Separate Controller Racks and Disks Racks New storage functionality Virtual Storage Machine CONTINUOUS Distributed Multi-Tenancy Virtual setting Resources Group Resources Group Resources Group System 1 System 2 System 3 VSP Virtual Storage Machine System 1 System 2 System 3 VSP G1000 Virtual disk controller which enables multi-tenancy and collaboration of distributed resources group across physical storage system boundaries N-Way Scale-Out EXTENSIBLE Dynamic and Intelligent Virtual Storage Machine Virtual Storage Machine (expansion) System Add new system Growth System 1 System 2 System 1 Virtual Storage Machine System 2 System 3 Virtual Storage Machine Load Balancing Data migration System 1 System 2 System 1 System 2 Non-Disruptive Migration CONTINUOUS Maintain application and data copies relationship Virtual Storage Machine SIMPLEX & PAIRED DEVICES System 1(Source) Data migration SIMPLEX & PAIRED DEVICES System 2(Target) • Support LDEV migration with replication works • Fault protection in the event of failure of Target storage during migration • Support SCSI-3 reservations • Possibility of NDA customers / partners and (not GSS) Hitachi Innovation: Global Storage Virtualization Global Storage Virtualization Create Virtual Storage Machines Abstract addresses from physical storage controllers global-active device Virtual LDEVs: 00:01 00:02 Virtual Storage Identity 12345 Virtual Storage Identity 12345 Resource Group 1 Resource Group 2 LDEVs ID: 10:00 10:01 10:01 10:02 10:02 LDEVs ID: 20:00 20:01 20:01 20:02 20:02 VSP G1000 S/N 12345 Hitachi VSP G1000 with Storage Virtualization Operating System Leverage Global-Active Device Functionality Deliver shared read-write volumes VSP G1000 S/N 23456 What is Global Active Device Synchronous remote copy of a data volume Active / Active configuration Campus Distance configuration a host is connected to both storage systems Metro Distance configuration a host has a primary path and a non-preferred path to insure the shortest path to the data is always used What is Global Active Device (cont.) Host sees a single volume on a single storage system ‒ Secondary storage system has a virtual storage machine (VSM) that looks like the primary storage system ‒ The primary and secondary volumes are assigned the same virtual LDEV number in the VSM ‒ Pair volumes are seen by the host as a single volume on a single storage system ‒ Both volumes receive the same data from the host Global Active Device Global Active Device achieves concurrent references/updates of mirrored volumes, while keeping data consistency Main DKC/Reserve DKC Prod. Server-1 (Active) App/ DBMS App/DBMS clustering Prod. Server-2 (Active) HA mirrored volumes for production servers App/ DBMS Mirrored volumes accept Read/Write I/O on both sides Requires 2 G1000 storage systems Quorum DKC HA mirroring Main DKC Reserve DKC QRM Quorum DKC Quorum disk, determines HA owner node in the case of a failure Any storage system can be used as long as it is supported by UVM Production Servers Clustered App/DBMS Active-Active processing with sharing the same data Global-Active Device with clustering management Storage Mgt Server(Active) HCS Prod. Server-1 (Active) HDvM Agent CCI App/ DBMS HCS offers efficient management of Global-Active Devices While providing central control of multiple DKCs App/DBMS clustering Storage Management Server Storage Mgt Server(Passive) HCS clustering HCS Prod. Server-2 (Active) HDvM Agent CCI App/ DBMS Pair Mgt Server Pair Mgt Server CMD HCS DB HA mirroring TC/HA mirroring MDKC CMD HCS DB RDKC QRM Quorum DKC Clustered HCS server is running, where the local HCS server enables HA management In the case of local site failure, the remote HCS server takes over the HA management HCS DB should be replicated with either TC or HA Pair Management Servers Runs HDvM Agent/CCI HCS management requests to configure/operate the HA mirrors via the command device GAD Failure Cases Failure Cases and Failback procedures ‒ Single path Failure ‒ Primary Storage System Failure ‒ All Active path to P-Vol fail for a Prod. Server-1 Prod. Server-2 single host in the cluster ‒ Quorum Disk Failure ‒ Storage Replication link failure ‒ WAN Storage Connection Failure ‒ Primary Site Failure ‒ Secondary Site Failure Primary Storage System Quorum Secondary Storage System System Configuration - Planning Data Centers Site Configuration 3 Data Centers Each storage system is located on the separate site Provides maximum level of business continuity for any type of storage system failures, or site failures (Local site, Remote site, Quorum site) Primary Storage System 2 Data Centers Primary Storage System and Quorum are located on the local site and Secondary Storage System is located on the remote site. Provides moderate level of business continuity for any type of storage system failures, or Remote site failure Secondary Storage System Primary Storage System Secondary Storage System A Single Data Center All the storage systems are located on the same site Provides business continuity for storage system failures, but cannot continue the business for a site failure Primary Storage System Secondary Storage System Quorum QRM Local site Quorum Remote site QRM Quorum site Remote site QRM Quorum Local site Local site New management functionality Hitachi Command Suite V8 Operational Efficiency Through Simplified, Integrated Management Hitachi Command Suite v8 STREAMLINED Management efficiency for the modern data center • Hitachi VSP G1000 support • HNAS support enhancements • Multipath management enhancements • 64-bit Architecture GLOBAL STORAGE VIRTUALIZATION REALIZED Create Virtual Storage Machine Global-Active Device Setting Up GAD – Set up global-active device dialog Launching “Set up global-active device” We have two launching points to invoke the landing page. - General Tasks in the Replication Tab - Shortcut in the [Actions] menu Add a launching point in General Tasks in the Replication Tab. Replication Tab Add shortcut in [Actions] [Actions] Menu Global-Active Device Implementation Service Global-Active Device Implementation Service from Hitachi Data Systems - Scope of the Service Implementation based on global-active device configuration: ‒ Consulting service to arrive at optimum solution design ‒ Integrate global-active device configuration into customer’s environment and validate the system recovery ‒ Configure global-active device, zoning for global-active device cluster, multipathing software, and production pairs ‒ Create design documentation, perform on a test server and demonstrate recovering from various paths and quorum array failures ‒ Provide best practices on platform to limited staff, demonstrate features, functionality, and knowledge transfer ‒ Provide project management and technical oversight Thank You