HA Has Never Been So Easy i Tera

“HA Has Never Been So Easy” i. Tera 5. 0 High Availability Common Europe – May 2007 Speaker - Andy Kowalski - Director Solution Services EMEA Vision Solutions Inc. [email protected]. com i. Tera HA - High Availability and real-time Disaster Recovery

Downtime Causes Of Downtime Solution Required Disaster Recovery High Availability Downtime refers to a period of time or a percentage of a timespan that a machine or system (usually a computer server) is offline or not functioning, usually as a result of either system failure (such as a crash or routine maintenance. Source: IBM HA Presentation, Eric Hess, April 2001 (Continuous Operations) * ‘HA’ generally refers to solutions that provide BOTH recovery and availability. Not all technologies provide a solution for BOTH…i. Tera 5. 0 HA does Reliability is not the same as Availability! i. Tera HA - High Availability and real-time Disaster Recovery 2

Cost Vs Recovery Time Cost i. Tera 5. 0 HA RTO = Minutes RPO = Last Transaction CDP =Continuous Data Protection / Point-in-Time Recovery Time Taken To Recover Cost depends on size of your business, the industry you are in, your relationship with customers and ability of your competitors to take advantage of the situation Time Taken To Recover is defined generally using RTO, RPO and CDP i. Tera HA - High Availability and real-time Disaster Recovery 3

i. Tera 5. 0 High Availability Business Benefits Technical Benefits Built-from the start for remote journaling CACH (concurrent audit concurrent heal ) E 2 KG (autonomic audit and fix) Replication all system critical data and objects User profiles, spool files, IBM job scheduler, device configurations, directory entries, IFS replication (journaling based) *, tape media library support (BRMS) * Apply accelerator (user configurable) Automated fail-over & virtual role-swap/rollover IBM clustering support * i. ASP support * MQ/Websphere support * Cluster host monitor (multi system/LPAR monitor) * * Option Easy to install Easy to monitor and manage Easy to support European & US support centers Remote implementations Costs less to buy & far less to own More flexible than HW mirroring Easier to use than competitors Highly scalable Highly dependable Extensive autonomic functions mean less time to manage i. Tera HA - High Availability and real-time Disaster Recovery 4 ‘High Availability Has Never Been so Easy, Affordable or Flexible’

Built-From-The-Start For Remote Journaling i. Tera 5. 0 HA Design Advantages i. Tera 5. 0 HA Built-From-The-Start For Remote Journaling Staging area is not needed Fewer jobs above MI interface Less I/O generated per transaction Data crosses MI Interface fewer times Easier to install Easier to manage & monitor More reliable (fewer moving parts) Faster delivery of data to backup DB Uses less disk resource Uses less CPU resource Other HA Solutions Have Remote Journaling Retro-Fitted Older HA technologies designed around local journaling which needs a staging area. RJ Retro-Fit was easier than full re-design i. Tera HA - High Availability and real-time Disaster Recovery ‘Retro-Fitted Is Not The Same As Built-From-The-Start’ 5

Comparison Of Data Transport Techniques i. Tera 5. 0 HA support RJ in async and sync mode Band- Comms Practical Over- Data Disables Width Interupt Distance head Loss Target Asynchronous Remote Journaling Low Synchronous Remote Journaling Med Traditional Local Journal Low Tolerant Intolerant Tolerant Data Integ. Select DASD Omit Req. High Low <1 sec No Transaction Yes Low High zero No Transaction High Med Hardware Mirroring Between i. ASPs High Intolerant Low Hardware Mirroring Between SANs High Intolerant Low High Low 30 sec zero No Transaction Any Yes Yes Disk sector Yes Any No i. ASP Disk sector No SAN Band-Width. How much band-width is required for mirroring data relative to other techniques Comms Interruptions. Is the transport technique tolerant or intolerant to these Practical Distance. Practically (not just lab) how far can it mirror relative to other techniques Over-head. How much does it impact performance of production relative to other techniques Data Loss. What is the max data loss in the event of an outage Disabled Target. Does it disable target database from being used (read-only) during mirroring Data Integrity maintained at business transaction level or at disk sector level Select/Omit. Does the technique allow libraries, objects, applications to be selected/omited DASD Req. Does it work with any storage or only with i. ASP or only with SAN i. Tera HA - High Availability and real-time Disaster Recovery Remote Journaling In Async Mode Offers Maximum Flexibility For HA 6

Apply Process Autonomic Heal Production (Source) System Backup (Target) System Record level heal needed transactions replicated Customer Master File (Replicated) Row update failed i. Tera 5. 0 HA – Design Advantages Eliminates the main cause of files ‘going on hold’ (row level update error) If a row update fails then file goes automatically into ‘heal’ mode while mirroring continues A concurrent job is automatically started which retrieves the failed record from the source file by relative record number (RRN) and repairs the record by RRN on the target file. Operators can review repair history at later stage One of Several Concurrent Helps to always maintain Role-Swap-Ready (RSR) status Audit Concurrent Heal (CACH) Features Makes HA easier to monitor and manage i. Tera HA - High Availability and real-time Disaster Recovery ‘Autonomicaly Heals Row Level Errors’ 7

Autonomic Apply Accelerator Production (Source) System Backup (Target) System Business Applications DB Files Latency Increasing decreasing Apply Job i. Tera 5. 0 HA – Design Advantages The Apply Accelerator helps to eliminate latency between remote journal and target DB If latency detected (user threashold) another apply route automatically and moves transactions in bulk using memory until latency eliminated Apply accelerator preserves FIFO integrity and can accelerate apply up to 400% Eliminates a time consuming operator activity Fixes latency issues effectively i. Tera HA - High Availability and real-time Disaster Recovery ‘Autonomicaly Heals Latency’ One of Several Concurrent Audit Concurrent Heal (CACH) Features 8

ZZ Audit Production (Source) System Backup (Target) System Business Applications Target Applications Users updating backup DB! ZZ Audit Heal DB Files Apply Job i. Tera 5. 0 HA – Design Advantages ZZ Audit looks for activity on backup that did not come from i. Tera HA and ‘heals’ it (rollback) This means that it is not essential to lock objects on the backup system during mirroring Improves performance Reduces potential for locking conflicts Improves usability of backup database i. Tera HA - High Availability and real-time Disaster Recovery One of Several Concurrent Audit Concurrent Heal (CACH) Features ‘Eliminates Need To Lock Backup Database’ 9

System Monitoring System Monitor shows all important status information for a production & backup system on a single screen E 2 KG Monitor runs continuously in the background and checks role-swapready (RSR) status at user defined intervals. Heals and notifies. Some tasks are continuous, Some daily and some weekly System Monitor Host Cluster Monitor allows multiple LPARs or systems to be monitored from a single screen E 2 KG (RSR) Monitor i. Tera HA - High Availability and real-time Disaster Recovery ‘Automatically Check That System Is Role-Swap-Ready’ 10

Replication Of All Critical Data & Objects Data (libraries, files) Data areas Data queues I FS (real-time journal based mirroring) MQ / Web. Sphere User profiles (real-time) Spool files & output queues Program objects Controllers, lines and devices IBM job scheduler entries Directory entries Tape media library support i. Tera HA - High Availability and real-time Disaster Recovery 11

Virtual Test Role Swap Run regularly (quarterly) to check backup system is role swap ready (RSR) Allows you to test startup procedures and business applications on backup system Click here for animation No effect on users on the production system of Virtual Test Role Swap No system exposure or data loss Can be executed in middle of the day Virtual Mode suspends the apply process on the target Replication continues and transactions are cached on the backup While in Virtual Mode processes can be executed on the backup node to test the applications including adding, updating data When tests are complete changes made on the backup are rolled-back and normal mirroring continues A log of each virtual test is produced Virtual Test Role Swap Initiation i. Tera HA - High Availability and real-time Disaster Recovery ‘Role-Swap Tests With No Downtime’ 12

Automated Role Swap Type *ROLLOVER is executed to allow the business to continue on the backup system while planned maintenance occurs on production system Role swaps are initiated from the backup system however, if your systems are clusterenabled, you can performed a role swap from any CRG and any system. Prior to the role swap type *ROLLOVER a series of automated pre-check are run by the E 2 KG monitor. This ensures that the backup is in Role Swap ready (RSR) status. Role Swap Type *FAILOVER is executed in the event of a unplanned production system outage and is therefore executed without pre-checks During the automated role swap a number of user configurable tasks are run in sequence. A monitor display a detailed view of status in real-time plus timings. Role-Swap Monitor i. Tera HA - High Availability and real-time Disaster Recovery 13

Popular Replication Scenarios 3 2 1 Production System (A) Backup System (B) Production System (A) One-to-One Replication Backup System (B) Production LPAR Backup LPAR Replicant/ Backup System (C) (A to B Scenario) One-to-Many Replication 5 Production LPAR Bi-Directional Replication (A to B & C Scenario) 4 Backup LPAR 6 Production LPAR (A) Production LPAR Backup system Production System Backup LPAR (B) Production LPAR Backup system Production System Backup LPAR Single-System Replication (A to B scenario) One-to-Many Replication i. Tera HA - High Availability and real-time Disaster Recovery Many-to-One Replication ‘Flexibility = Business Value’ 14

Other Important Features - 1 No downtime during install i. Tera HA is designed so that the implementation and initial synchronization of the backup system can be done without introducing any user down-time. Auto Installation When i. Tera HA is loaded on the source system it sends an copy of itself to all target nodes and auto installs. Journals are automatically created. Library Analyzer i. Tera HA analyzes all libraries and objects and provides valuable input to help optimize the mirroring setup. Leverages IBM PTF Process Already familiar to most System i operators. i. Tera HA PTF’s can be downloaded from a Vision FTP site directly to the System i, cover letters can be reviewed for each PTF and the PTFs can be temporarily or permanently applied while HA remains active. i. Tera HA - High Availability and real-time Disaster Recovery 15

Other Important Features - 2 Auto Library & Object Registration. Customer who periodically create rename and delete libraries and objects and want them automatically mirrored can do so. Libraries & contents can be defined for mirroring with wild cards. Libraries and contents will be automatically journaled, synchronized and mirrored. Library Auto Sync. Helps eliminate the need for operators to be present during the initial data synchronization. Libraries to be replicated are pre-selected, the journals to which the libraries will be replicated are assigned, and a syncing priority status is assigned. Sophisticated targeting functionality allows for a wide range of unattended syncing possibilities. MQ / Web. Sphere Support. i. Tera HA replicates the associated MQ Library. Creates the Queue Manger, if needed. Creates the queue. Replicates the current messages and other IFS objects. Creates the remote journal and receive. Applies the messages to the Backup node, then deletes messages as needed. i. Tera HA - High Availability and real-time Disaster Recovery 16

Other Important Features - 3 Easy to use Menu Fastpaths All menus are logically and consistently laid out following i 5/OS naming conventions and standards to make them familiar and easy to use. In addition a fastpath technique is used to allow operators to navigate faster and communicate easier to others. To access the system monitor for example type ‘ 1. 1’ on any screen…it takes you to menu 1 and then option 1 in a single step. Triggers i. Tera HA executes a trigger search and keeps a record of all triggers in the mirrored database. It is possible to specify how the trigger should be handled during Role Swaps (*ROLLOVER/*FAILOVER). Triggers are generally disabled during mirroring to prevent duplication of updates and enabled when in *ROLLOVER/*FAILOVER MODE. Constraints i. Tera HA executes a constraint search and keeps a record of all constraints found in the mirrored database. It is possible to specify how the constraint should be handled during Role Swaps (*ROLLOVER/*FAILOVER). Constrains can be in enabled of disables state. i. Tera HA - High Availability and real-time Disaster Recovery 17

Other Important Features - 4 Audit Command Console The Audit Command Console is a central location from which to monitor and maintain all of the i. Tera HA audits. Each audit (there are many) is defined with a minimum interval time frame in which, if the audit hasn’t run, the Audit Command Console runs it. CRC Audit Allows all records in a file to be checked between source and target without sending the records to the target. A CRC code is generated for each record (or block of records) on source and target and only the CRC code needs to be compared. This greatly reduces the time involved. IBM Clustering Support i. Tera HA is fully cluster designed and works with IBM Cluster services (recommended optional). Independent Auxilary Storage Pool (i. ASPs) i. Tera HA can be installed in i. ASP environments. i. Tera HA - High Availability and real-time Disaster Recovery 18

Easy To Install, Monitor & Manage Other Vendor’s i. Tera HA Installation Typically 120+ Man-Hours Installation Typically 20 -40 Man-Hours Monitoring Typically 5 minutes per day Role Swap Tests Virtual test role swaps can run at any time. Easy to schedule because introduces no down-time Monitoring Typically 30+ minutes per day Role Swap Tests Days of planning & execution Introduces down-time i. Tera HA - High Availability and real-time Disaster Recovery ‘It’s The Next Generation of HA’ 19

24 X 7 Support Vision have 3 support centers which allows full 24 X 7 support for European customers (not just an on-call person). Support From Manchester UK Support From Los Angeles USA Support From Salt Lake City USA i. Tera HA - High Availability and real-time Disaster Recovery ‘Follow the Sun Support’ 20

Other Important Business Benefits - 1 Costs Less To Buy List prices are lower than competitors products and are based on power of your System i servers. Costs Far Less To Own Costs less to install and a lot less to manage than competitors solutions. No Need For a ‘Lite’ Version i. Tera HA product features are not restricted for example there is no restriction on the number of journals/links/CRGs. i. Tera HA is suitable for all customers SMB and Enterprise. No Customization Needed No need for customization. The same product supports all business applications. i. Tera HA - High Availability and real-time Disaster Recovery 21

Other Important Business Benefits - 2 Highly Reliable No legacy code, newer more efficient design with fewer moving parts make i. Tera HA highly reliable. This is not a niche product, it’s a global product with over 600 implementations. Local Expertise Available There are over 50 Vision partners with implementation skills in EMEA Very High Transaction Volume. High performance data replication possible because there is no data staging and no legacy code. Efficient use of remote journaling and memory based accelerator Is typically more effective for performance than multiple apply jobs which add CPU and I/O. This is particularly effective on Power 5 CPUs. Extreme high workload environments may require multiple journals. This makes i. Tera HA highly scalable. i. Tera HA - High Availability and real-time Disaster Recovery 22

Customer Comments ‘In just 2 -1/2 weeks we went from installation to (our first) successful rollover’ Darrell Martel, Director Server Technology, Target Corporation ‘Echo 2 let’s me sleep on weekends’ Dan Johnson. Sr. Mgr. Large & Midrange Systems, Fedex Trade Networks ‘Echo 2 practically runs itself’ Tony Pack, Database Administrator, The Dixie Group We regularly and successfully test role swaps with Echo 2’ Jim Ravens, Business Recovery Manager, Boyd Gaming Empire Corporation ‘We had a power problem…the failover worked great! We ran production for about 2 hours- and then failed back!’ Dean Choudri, IS Manager (i. Series), Empire Corporation i. Tera HA - High Availability and real-time Disaster Recovery ‘Our Customers Love it !’ 23

Competitive Replacements 30% of i. Tera HA customers are companies who replaced another vendor’s product because of failover issues, out of sync issues, time management issues & cost of ownership issues. MIMIX Replacements Ahlstrom Fiber Advance Auto Arnold Logistics Automated Installment Bank Intesa Bank Colpatria BSG Sauder Woodworking Campbell Soup Chapman Auto Chisago County Electrolux Husqvarna Fed. Ex Healthnet ICON Health and Fitness LDS Distribution Lease. Plan MBS Textbook Nationwide Insurance North American Stainless MIMIX Replacements Nucor Steel Texas Nucor Steel Arkansas Nucor Steel South Carolina Perkins Paper RG Financial Resorts Casino Seneca Gaming Simonds International Standard Textile State Univ. Retirement Systems Time Warner Cable Tree of Life Trim Masters Turning Stone Casino Verture Memberworks Wabash National Washington County Woodside Homes Wells Fargo DATAMIRROR & NOMAX Replacements Alaska Native Medical Center Ambrake Ash City Big 5 Sporting Goods City of Sheboygan Columbia Insurance Dot Foods Discovery Channel Communications DME Evans and Wood Federal Home Loan Bank First Advantage Great Plains Communication Hitachi Automotive Products Meridian Automotive Mesilla Valley Oxford Automotive RLI Insurance Company Stryker Insurance Medical Plus many more… i. Tera HA - High Availability and real-time Disaster Recovery 24 ’ 30% Of i. Tera HA Customers Replaced Another Vendor’s Product’

Over 600 Customers Already In Production i. Tera HA - High Availability and real-time Disaster Recovery ‘The Most Sold HA Solution’ - IBM 25