IBM e Server i Series Session 36 CE
IBM e. Server i. Series Session: 36 CE 410253 i. Series HA and Storage Solutions Steven Finnes finnes@us. ibm. com www. ibm. com/eserver/iseries/ha 8 Copyright IBM Corporation, 2003. All Rights Reserved. This publication may refer to products that are not currently available in your country. IBM makes no commitment to make available any products referred to herein.
IBM e. Server i. Series Agenda § 2003 i. Series Storage Solutions §New integrated disk attach: PRICE/PERFORMANCE !!! §i. Series and SAN §Extending i. Series connectivity to Enterprise Storage Server (ESS) §i. Series Availability Solutions §Positioning §Support, Planning and Education §Summary © 2003 IBM Corporation
IBM e. Server i. Series Disk Solutions i. SCSI Integrated SAN NAS TCP/IP SCSI TCP/IP over LAN SCSI over LAN SAN i. Series Integrated Storage Small- to Enterprise. Sized Environments Enterprise Storage Server (ESS) Network Attached Storage i. SCSI Medium- to Enterprise. Sized Environments Can be connected for i. Series Integrated File System (IFS) serving (via NFS) Not Supported © 2003 IBM Corporation
IBM e. Server i. Series NEW: High Performance Integrated Storage §Achieve up to 3 X throughput improvements with enhanced PCI-X I/O options for integrated disk attach §Improve performance with IBM’s 3 disk optimized RAID-5 §New 35 GB and 70 GB 15 K RPM disk drives §New PCI-X I/O towers with rack mount options © 2003 IBM Corporation
IBM e. Server i. Series Integrated Scalability and Performance §Exploits 1 Gb/s High Speed Link (HSL) connectivity §Scales from 8 GB to 144 TB ƒ High performance PCI RAID Disk Unit Controller supporting up to – 2047 internal disk units §Aggregate transfer rate of 640 MB/s on each controller § 235 MB distributed array level write cache ƒ Up to 106 GB compressed write cache total Ÿ 757 MB per 15 disk units CPU, L 1, L 2, L 3 Cache Main Storage and Expert Cache HSL Bus, IOP CTLR -757 MB Write cache 4 x 160 MB/s Bus Up to 18 drives per IOA §Ultra-3 SCSI and 15, 000 RPM disk units §Hardware assisted and optimized RAID-5 NOTE: Ships with Ultra-3 support © 2003 IBM Corporation
IBM e. Server i. Series New Ultra* SCSI Disk Adapters §#2757 High Performance RAID-5 §#5705 Tape/Disk Adapter ƒ Non-RAID-5 ƒ 235 MB write cache, up to 757 MB ƒ Internal Ultra SCSI bus with compression –Supports up to 6 Disk Units and ƒ RAID-5 optimized, minimum 3 internal optical / tape device drives in array ƒ External Ultra SCSI bus ƒ 4 Ultra SCSI buses –Optional tape devices capable of ƒ Supports up to 18 disk units and attaching to LVD SCSI two internal optical / tape devices §#2782 Entry RAID-5 Adapter ƒ 40 MB write cache ƒ RAID 5 optimized, minimum 3 drives per array ƒ Two Ultra SCSI buses *NOTE: Shipped today with Ultra 3 (160 ƒ Supports up to 12 disk units and mb/s) support. two internal optical / tape devices © 2003 IBM Corporation
IBM e. Server i. Series High Performance RAID Controller Comparisons Feature 2778 / 4778 2757 Improvements SCSI bus 80 MB/s 160 MB/s* 2 x faster # SCSI buses 3 4 1. 25 x more Max PCI Burst Rate 133 MB/s 532 MB/s 4 x more Processor Speed 80 MHz 500 MHz 6. 25 x faster Compressed Write Cache 104 MB 757 MB 7 x larger Min/Max drives in RAID 5 array 4 / 10 disks 3 / 18 disks Optimized SCSI bus tagged command queuing N/A Yes Faster Response Time (under heavy I/O load) Array parity checking and memory scrubbing Yes - New HDW assist 5 x faster RAID Configuration Enable or Disable Capacity, Performance, Balance or disable Greater Flexibility *NOTE: Shipped today with Ultra 3 (160 mb/s) support. © 2003 IBM Corporation
IBM e. Server i. Series © 2003 IBM Corporation
IBM e. Server i. Series Storage Specialist ? ( Truth be known. . . there is no such thing !) Virtualized integrated storage means not much to do here. . © 2003 IBM Corporation
IBM e. Server i. Series Integrated Storage Management §Self-caching, automatic load balancing ƒ Data is automatically spread across all disk units in a Disk Pool ƒ Parallel I/O results in optimum performance and disk utilization ƒ No user planning or management, no manual data placement ƒ No individual "disk full" conditions to handle ƒ Integrated tools to study effects of read cache before implementation §Added disk capacity is automatically utilized §Data is spread by OS/400 Storage Management, not disk controllers ƒ Regardless of physical attachment solution © 2003 IBM Corporation Object A Object C Object B OS/400 TIMI Storage Management I/O Object D
IBM e. Server i. Series Storage Maintenance Interface - Single Point of Control §Create, manage and monitor storage virtualization on i. Series ƒ Secured access ƒ Configuration, protection, availability, recovery and maintenance §Self-guided wizards §Graphical representation of complex management tasks §SAN-like storage management for Windows servers © 2003 IBM Corporation
IBM e. Server i. Series V 5 R 2: Multiple independent databases, system libraries, and library names Auxiliary Storage Pools §Independent Disk Pools ƒ Required for switched disk clustering ƒ New option for single server and redundant server availability ƒ Maintenance by Independent Disk Pools §Outage isolation ƒ Server and other ASPs continue if an IASP fails §Application consolidation ƒ Up to 223 IASPs with V 5 R 2 User ASP (journal receivers and save files) User ASP (libraries, UDFS) © 2003 IBM Corporation Independent ASP Company 1 (libraries, UDFS) System ASP Independent ASP Company 2 (libraries, UDFS) Independent ASP Payroll (libraries, UDFS)
IBM e. Server i. Series HABP Switchable IASP Cluster for Websphere Database Application Server Firewall Java Clients WAS 1 Http Server Application Server Admin Server Clone Journaling Active WAS 2 Web Clients Http Server Application Server Admin Server Switchable IASP Admin Repository Persistent Servlet Session Data Clone Admin Repository Redundant firewall, application, Switchable Disk Pool © 2003 IBM Corporation Enterprise Data
IBM e. Server i. Series HA Replication/Clustering Solution for Websphere Database Application Server Firewall Primary DB Java Clients WAS 1 Http Server Enterprise Data Application Server Admin Server Clone Admin Repository WAS 2 Persistent Servlet Session Data Sync RJ Secondary DB Web Clients Http Server Application Server Admin Server Clone Redundant firewall, application and DB © 2003 IBM Corporation Admin Repository Enterprise Data Persistent Servlet Session Data
IBM e. Server i. Series SAP with Independent ASP's §Data Center Topology ƒ IASPs for multiple DB images ƒ Switchable IASPs for availability © 2003 IBM Corporation
IBM e. Server i. Series © 2003 IBM Corporation
IBM e. Server i. Series Storage Virtualization for Windows and Linux Servers §i. Series storage consolidation for multiple environments §Extends SAN-like functions to: ƒ Windows Servers –Virtual drives per server, up to 2 TB –Dynamically increase disk size for Windows 2000 servers –Unique hot spare facility provides simple, efficient high availability ƒ Linux Servers –Storage Spaces are 1 MB to 64 GB each –Up to 20 Storage spaces per Linux partition §Improves performance with more disk arms §Consolidated backup of OS/400, Linux and Windows Servers §Graphical management © 2003 IBM Corporation OS/400 Disk #1 Disk #2 Disk #3 Integrated x. Series Server Integrated x. Series Adapter
IBM e. Server i. Series OS/400 i. Series Storage Management Options: Integrated and SAN Storage Management IOA Card (FC) IOA Card All managed by OS/400 SAN Storage Solution SAN Storage Software Tools Customer managed hubs/Switches, Hubs, Routers, etc. Host Bay External Storage HBA SAN Software Tools Host Bay HBA Processor SSA Adp. Processor Cache SSA Loop Cache NVS © 2003 IBM Corporation Integrated Storage Solution NVS ** recall that multipath is not supported yet
IBM e. Server i. Series Enterprise Storage Server §Family of enterprise storage servers from entry through Model 800 with Turbo feature ƒ Increased throughput (compared to F 20) ƒ 15 K RPM disks §Built for Redundancy § 2 GB Fibre Channel/FICON § 64 GB Non-volatile cache ƒ 2 GB fixed write cache §Data protection ƒ RAID-5 ƒ RAID-10 §Flash. Copy, PPRC © 2003 IBM Corporation
IBM e. Server i. Series Enterprise Storage Server - Stor. Watch §Comprehensive ESS Management Tools ƒ Web-based management ƒ One or more ESSs - any location §Stor. Watch ESS Specialist - built-in ƒ Status, configuration, authorization ƒ Notification (e-mail, pager, SNMP) ƒ Capacity reassignment ƒ Copy Services management §Stor. Watch ESS Expert ƒ Optional ƒ Performance, asset, capacity management © 2003 IBM Corporation
IBM e. Server i. Series Fibre Channel Tape Connectivity §Connectivity through #2765 Fibre Channel Adapter ƒ Improved performance, up to 136 GB/hr* per drive 1. 3 TB /hr with 10 drives ƒ Maximum 2 FC adapters per IOP §Enhanced connectivity and resource sharing §Multiple targets, 16 devices per adapter ƒ Share tape resource §Maximum distances with 2 GB support *Note: Save rates based on OS/400 V 5 R 1 on i. Series Model 840 and IBM 3590 -Exx using SCSI and #2765 Fibre channel adapter. © 2003 IBM Corporation
IBM e. Server i. Series and SAN Options with V 5 R 2 Switch fabric expands options across multiple initiators and multiple targets §Enabled by IBM 2109 switch Multiple SAN targets can be connected through a single i. Series fibre channel adapter §Optimizes number of fibre channel adapters §Zoning recommended for performance FC adapter for Linux partition Switched disk support © 2003 IBM Corporation IBM 2109
IBM e. Server i. Series Disk Attach Solution Guide File Serving OLTP Dedicated environment Dedicated ESS disks Integrated i. Series storage for Windows Servers, LINUX, OS/400 PASE, Domino Applications i. Series direct-attach disks §Interactive, I/O intensive commercial transaction processing §Batch processing dependent on disk response time §Predictive response times required at all times © 2003 IBM Corporation SAN environment SAN attached ESS disks §File serving, compute-intensive workload with large block I/O §Web serving (downloading Web pages), Domino / Lotus Notes, other large block I/O applications
IBM e. Server i. Series Availability Solutions © 2003 IBM Corporation
IBM e. Server i. Series The HA Market: ŸTop business issue for IT executives ŸDemand for HA solutions has been growing at 20% CGR for the past three years Ÿe Infrastructure is driving the demand for 24 x 365 operations Customer: ŸServer Consolidation ŸBanking Srvcs, Financial Srvcs, Telecom, Manufacturing, e. Commerce , Health Systems, ŸDistribution/Retail, Insurance, Health, Transportation Value: ŸEliminate planned down time = asset utilization Ÿback-up window + maintenance = $$ ŸMinimize unplanned down time = business continuity ŸLocal unplanned scenarios ŸDisaster recovery scenarios Source: IMEX Research. com Jan 2002 @ http: //www. highavailabilitycenter. com/index. shtml © 2003 IBM Corporation
IBM e. Server i. Series Define Business Objectives Define the outage types Business Impact Analysis Disruption impact Risk Analysis Probability of disruption critical function SLAs Service Level Agreements RPO Recovery Point Objective 24 hrs, 1 hr, immediate ? RTO Recovery Time Objective 6 hrs, 1 hr, 10 min ? © 2003 IBM Corporation Scheduled outages The majority of all downtime is planned Backups PTF and OS installs Application maintenance Upgrades Unscheduled outages Locally Recoverable Outage Application failure Operator error Power outages Network failure Hardware failure Disaster Loss of the IT facilities
IBM e. Server i. Series Establish Requirements §Understand customer availability needs and expectations: §What outage types ƒ Planned? ƒ Unplanned? ƒ Disaster Recovery? ƒ All of the above? §High Availability (HA) or Continuous Availability (CA) requirements? ƒ Near zero downtime for any outage type? ƒ Loss of transactions not tolerated? §Availability level vs. . cost vs. . complexity ƒ Complexity varies, but all solutions creates additional complexity §Long distances between the primary and backup servers? ƒ Communication costs © 2003 IBM Corporation §Availability discussions should include ƒ 1. Server hardware ƒ 2. Storage (disk and tape storage) ƒ 3. Software and Automation ƒ 4. Networking ƒ 5. Services Determine the cost vs. . Recovery Point Objective and Recovery Time Objective
IBM e. Server i. Series Translate to business metrics Outage Type Current Duration Cost to Business Objective Planned 50 hrs $200 K 0 hrs Unplanned 6 hrs $500 K 1/2 hrs Disaster 48 hrs $2 M 12 hrs §Total the annual Planned Outage time §Estimate an Unplanned Outage Risk §Estimate Cost/Value to the Business §Recommend Appropriate Solutions §Consider Using an ROI tool © 2003 IBM Corporation
IBM e. Server i. Series Evaluate Solution Strategies §Replication Based Solutions §Switchable Storage Pools (IASPs) §Clustering §SAN-based Methods © 2003 IBM Corporation
IBM e. Server i. Series Availability Tools Save While Active HA Software Copy Services SWA ŸIncluded in OS/400 (no additional charge) ŸUsed for Backups only Ÿ 5 -15 minute outage for checkpoint ŸFairly Easy ŸUsed for HA/DR or Tape backups or planned outages (e. g. release upgrades) ŸNo outage ŸAll components are protected ŸLonger distances are possible ŸSecondary System can run other applications on back-up data ŸIntegrity at Application Level Ÿmost comprehensive and widely used ** SWA and HA work on internal or external disk ** © 2003 IBM Corporation ŸUsed for DR or Tape backup (no help for release upgrades) Ÿ 30 -60 minute outage to make Flash. Copy Ÿmulti-hour outage at DR time ŸAll components can be protected ŸSecondary System stops other work at DR/Backup time ŸIntegrity at byte-level only ŸProcedure is Fairly Easy (repetitive) although care is required
IBM e. Server i. Series Comprehensive i. Series HA Solution Suites Tape Data & CPU concurrently useable on backup server Primary Server Backup Server - active copy accessible in real-time 1000 s of customer accounts WW ŸCan provide the most resilient availability solution ŸScheduled Outages: Concurrent Saves & Maintenance Procedures ŸUnscheduled Outages: Failover support for primary server failure ŸWorkload balancing (Queries, Batch, Web etc. ) on Backup server ŸNo IPL required for switchover © 2003 IBM Corporation
IBM e. Server i. Series Classic HA Solution Concepts §HA best case RTO/RPO ƒ Sync Remote Journaling should be considered ƒ Switched disks may play a role ƒ HSL/LAN/WAN Connectivity §Determine solution objectives §Use LPAR to help gain maximum ROI §Use Clustering to help gain maximum availability © 2003 IBM Corporation
IBM e. Server i. Series Server Consolidation & High Availability Replication vs. IASP topologies Shrink Development Partition at switch-over time HA based on replication - second copy "live" - back-up & data access © 2003 IBM Corporation HA based on switched disk - no second copy - no data access until switch
IBM e. Server i. Series The Service Bureau Approach §HA or DR strategy ƒ Distributed sites backed- up centrally © 2003 IBM Corporation
IBM e. Server i. Series Assess solution functional attributes. . §Replication validation tools and functionality (ready to switch) §Recovery Automation §IFS, Data Areas, Data Queues §IASP integration and support §MQ support, Websphere support §Centralized availability solutions for heterogeneous environments §Application level granularity for availability management §Provides both switch over and switch back capability §Can it replicate all object types §Advanced auditing and verification function §Out of sequence events in journal entries §Automated Object Replication management (creates, deletes, moves, renames) §Integrated Remote Journaling §Journal Minimal Changes §Journal Standby Mode §Large Object Support © 2003 IBM Corporation §Integrated Clustering Middleware §Can demonstrate Cluster. Proven application support §OS/400/Linux/Windows integrated solution
IBM e. Server i. Series Assess solution provider attributes and track record §Usability: Ask for customer reference accounts that role swap once a month §Usability: Speak to a customer reference that has role swapped during an unplanned outage §Reference accounts: what are they doing, how did they implement, how are they using it §End-to-end implementation plans §Strong local partner practice § 24 x&7 International and local service & support infrastructure §Diverse service offerings §Support Publications © 2003 IBM Corporation §Full spectrum of offerings §Solutions spanning simple DR to continuous availability §Ease of Use/Systems management capabilities §WW presence - large install base §WW language support - indication of company capability, strength and experience §Strategic development relationship with IBM §Industry solution provider relationships
IBM e. Server i. Series Assess Total solution attributes. . §"Half" of an HA solution is service & support §Investigate customer reference accounts on both implementation and post implementation service and support §If you are using or deploying a package like SAP, JDE, Intensia. . . find out how many they have done, how many are in production, talk to a reference account on how the project went §Does the solution provider support Cluster. Proven ISV solutions §Project Plans (ask for a project plan documentation example) §Project Ownership, Project Manager, Implementation Plan §Test Plan, Training Plan, §What type of technical training certification is required for the implementation experts in both the ISV solution and i. Series §Post Installation Support, how many support people per installed customer ? Training and certification © 2003 IBM Corporation
IBM e. Server i. Series Positioning Availability Solutions §i. Series Clustering and High Availability Business Partner Solutions ƒ Most widely deployed type of HA solution across i. Series install base ƒ Can be the most resilient and diverse coverage model ƒ Ask reference accounts about service and support §Single server and Save-While-Active (SWA) for backups §For disaster recovery ƒ i. Series Replication Clusters (internal or ESS storage) ƒ OS/400 Mirrored ESS ƒ Hosted i. Series Services or Business Continuity Recovery Services §For basic planned and unplanned outages ƒ i. Series Switched Disk Clusters ƒ Best solution will often be a combination of one or more availability technologies §Analyze your availability requirements first ƒ Engage your i. Series rep or HA expert for assistance © 2003 IBM Corporation
IBM e. Server i. Series Trademarks and Disclaimers 8 IBM Corporation 1994 -2003. All rights reserved. References in this document to IBM products or services do not imply that IBM intends to make them available in every country. The following terms are trademarks or registered trademarks of International Business Machines Corporation in the United States, other countries, or both: AS/400 e e. Server IBM (logo) i. Series OS/400 Lotus and Smart. Suite are trademarks of Lotus Development Corporation and/or IBM Corporation in the United States, other countries, or both. MMX, Pentium, and Pro. Share trademarks or registered trademarks of Intel Corporation in the United States, other countries, or both. Microsoft and Windows NT are registered trademarks of Microsoft Corporation in the United States, other countries, or both. Java and all Java-based trademarks are trademarks of Sun Microsystems, Inc. in the United States, other countries, or both. SET and the SET Logo are trademarks owned by SET Secure Electronic Transaction LLC. C-bus is a trademark of Corollary, Inc. in the United States, other countries, or both. UNIX is a registered trademark of The Open Group in the United States and other countries. Other company, product or service names may be trademarks or service marks of others. Information is provided "AS IS" without warranty of any kind. All customer examples described are presented as illustrations of how those customers have used IBM products and the results they may have achieved. Actual environmental costs and performance characteristics may vary by customer. Information in this presentation concerning non-IBM products was obtained from a supplier of these products, published announcement material, or other publicly available sources and does not constitute an endorsement of such products by IBM. Sources for non-IBM list prices and performance numbers are taken from publicly available information, including vendor announcements and vendor worldwide homepages. IBM has not tested these products and cannot confirm the accuracy of performance, capability, or any other claims related to non-IBM products. Questions on the capability of non-IBM products should be addressed to the supplier of those products. All statements regarding IBM future direction and intent are subject to change or withdrawal without notice, and represent goals and objectives only. Contact your local IBM office or IBM authorized reseller for the full text of the specific Statement of Direction. Some information in this presentation addresses anticipated future capabilities. Such information is not intended as a definitive statement of a commitment to specific levels of performance, function or delivery schedules with respect to any future products. Such commitments are only made in IBM product announcements. The information is presented here to communicate IBM's current investment and development activities as a good faith effort to help with our customers' future planning. Performance is based on measurements and projections using standard IBM benchmarks in a controlled environment. The actual throughput or performance that any user will experience will vary depending upon considerations such as the amount of multiprogramming in the user's job stream, the I/O configuration, the storage configuration, and the workload processed. Therefore, no assurance can be given that an individual user will achieve throughput or performance improvements equivalent to the ratios stated here. Photographs shown are of engineering prototypes. Changes may be incorporated in production models. © 2003 IBM Corporation
IBM e. Server i. Series Questions? Submit your questions now by clicking on on the “Ask A Question” button in the left corner of your presentation screen. Steve will answer your questions off-line. The questions and answers will then be posted on Search 400. com. We will e-mail you when your question has been answered. © 2003 IBM Corporation
- Slides: 40