Technical Program

Business Continuity

Sessions focus on ways to make HP NonStop systems even more available. Presentations which might have this focus include those which address: ways of designing applications and configuring systems for availability; operational and system management procedures; and business continuity procedures and experiences.

BC-1457-U
Batch Processing for Disaster Recovery/Business Resumption on the NSK
Ries Wytenburg, Independent Contactor
Audience Level: Intermediate

Nearly all NSK users have some component of there business processing that is considered “batch”. How does one implement a batch processing environment that will provide optimal production control and progress reporting under normal processing conditions as well as production failover to the disaster recovery/business resumption systems and comply with all the security and operations policies and regulations?

This presentation talks about a real and live “whole” system way to implement batch processing utilizing standard NSK utilities in a way that provides effective operability while ensuring the backup/failover system batch environment is up to date and disaster provides recovery check-points for re-start processing on the fail-over system. In fact the batch environment on both primary and the backup is actually (partially) a live-live configuration.

BC-1458-U
Online DB Synchronization in an RDF Environment for Disaster Recovery
Robert Sardiga, DataForce and Ries Wytenburg, Independent Contactor
Audience Level: Intermediate

Every Remote Database Facility ( RDF) user has a requirement to synchronize their database to the disaster recovery system at least once. The procedures documented in the RDF manual require some application outage, an application outage that is not practical for most on-line applications. Backup cannot handle open files/tables. TMF cannot easily be used to restore data to another system.

SyncWorks is an architecture and a set of procedures and routines that enables the creation or recreation of all or part of the backup database on the D/R system without stopping the production application. The procedure uses standard NSK utilities to reliably move hundreds of tables and gigabytes of data while the primary application continues to handle on-line transactions.

This presentation details a real live implementation of an on-line database synchronization that utilizes standard NSK products and utilities to provide a comprehensive operational environment. The approach is systematic, repeatable and controllable for a complete or partial database resynchronization and requires no application outage.

BC-1641-U
Go Active/Active for your mission critical HP NonStop Applications & Achieve Continuous Availability During Planned and Unplanned Outages – It’s Good for The Business and It’s Good for IT
Chris McAllister, GoldenGate
Audience Level: Intermediate

Achieve continuous availability during planned and unplanned outages – it’s good for the business and it’s good for IT. This session is an interactive journey through the process of evaluating your application and designing a strategy for moving your organization to a more highly available environment for all of your operationally critical information. This session investigates the pros and cons of the myriad replication technologies and philosophies that all culminate at the ultimate high availability topology, Active/Active. Learn about physical vs. logical and synchronous vs. asynchronous replication; process to process vs. decoupled architecture; conflict detection algorithms and conflict resolution best practices. Along the way we will discuss how to keep the lights on during migrations and upgrades (O/S, DB, and Applications).

BC-1802-HP
NonStop Server Business Continuity Update
Robert Loftis and Jim Willis, HP
Audience Level: Introductory

See what’s new, or coming soon, to the full product set for HP’s NonStop disaster tolerance solution. You’ll hear about plans underway for the release of RDF 1.9 for HP Integrity NonStop and HP NonStop Blade servers, as well as improvements, such as “transaction typing,” coming for NonStop TMF. As expected, this year’s enhancements for AutoTMF and AutoSYNC will be outlined, and you’ll get a preview of a brand new complement coming soon for RDF: HP NonStop SQL DDL Replicator, or SDR for short. SDR will automate the replication of Data Definition Language operation changes to RDF’s target system(s) with the kind of manageability, power, data integrity and performance you have come to predict from HP NonStop products. We’ll also review the latest HP ODBC MX on OSS release, and a new product from HP to help partners accomplish new active/active replication solutions.

BC-1803-HP
New HP Solution for Replicating NonStop SQL DDL: SDR
Richard Carr, Carr Scott Software Inc. and Robert Loftis, HP
Audience Level: Intermediate

Join us for an introduction to a new member of the HP NonStop Business Continuity suite of products: NonStop SQL DDL Replicator. SDR, as it is known, automates Data Definition Language (DDL) replication for NonStop SQL MP databases on S-Series and Integrity NonStop servers that use the NonStop Remote Database Facility (RDF). SDR is being developed based on HP NonStop customer requirements and the first shipment is targeted for late summer, 2008, focused on SQL MP.

SDR captures, replicates and implements DDL operations not currently replicated by RDF or 3rd party software. It works to ensure that data and DDL replication are in the correct sequence to guarantee logical consistency. It also requires no changes to application programs and causes no significant effect on performance. SDR does not impact the operation of primary system applications or transmission of database updates to backup systems. As expected, SDR provides low latency in keeping the backup database in sync to ensure a rapid and precise takeover when needed.

BC-1917-U
Active/Active 101
Bill Highleyman, The Availability Digest
Audience Level: Intermediate

Active/active systems can achieve extreme availabilities of six, seven, or even eight nines (literally, only seconds of downtime per year). An active/active system is a network of independent processing nodes actively cooperating in a common application. Should a node fail, all that needs to be done is to switch over that node’s users to a surviving node. Recovery is in subseconds to seconds. Though active/active systems have been successfully deployed in the NonStop community for several years, the technology is applicable to OpenVMS and HP-UX systems as well.

Active/active systems can provide century MTBFs, can recover from any failure in subseconds to seconds, can eliminate planned downtime, can provide disaster tolerance, can enable application scaling, can provide risk-free failover testing, can use all available capacity, and can be load-balanced at will. Best of all, these benefits can often be achieved with little or no extra cost.

This presentation explores how to configure active/active systems and explains why such high availabilities are achieved. The underlying infrastructure required to implement such systems is discussed. The many other advantages of active/active systems are described. Several contemporary examples of successful active/active systems in production today are presented.

BC-1927-U
Advantages of Multinode Active/Active Architectures
Bill Highleyman, The Availability Digest
Audience Level: Intermediate

Much has been said recently about implementing systems as a set of cooperating nodes in an active/active configuration, and many such systems are now in production. Such configurations can achieve uptimes measured in centuries - in effect, virtually 100% uptime.

However, there are many additional benefits of a multinode architecture. These include:

Extremely high availabilities (six 9s and beyond)
Extraordinary recovery times from a failure, measured in seconds
Disaster tolerance
Elimination of scheduled downtime
Data locality
Efficient use of all purchased capacity (no standby system)
Simplified, risk-free, and fast failover testing
Application scaling

This paper explores these advantages of multinode configurations. It also delves into the economics of these systems in terms of total cost of ownership (TCO) as compared to monolithic architectures (i.e., the costs of multinode systems compared to monolithic systems).

BC-2417-U
Case Study in a Multi-site Migration from S-series to Integrity NonStop while Integrating a Multi-site Virtual Tape Library System while Protecting the Legacy Assets
Mike Allen, Verizon
Audience Level: Introductory

Migrating a multi-site, multi-customer, and multi platform backup infrastructure from a legacy tape backup solution to a fully automated, secure Virtual Tape Library System. The effort presented a multitude of technical challenges which can be categorized under the descriptors of application/processor, storage, and communications integration. This presentation will describe for each category the technical considerations that went into the planning, the differences encountered between the plans and the realities of implementation (including how the business goals changed during the overall effort), and comment on the overall successes and lessons learned from the effort.

BC-2418-U
Active/Active in the Real World
Neil Barnes, LloydsTSB
Audience Level: Introductory

We have been running an Active/Active environment on our Tandem systems for over 10 Years. Active/Active is seen as a way of improving availability, this session will cover it being used in a live environment. The highlights of the presentation will cover:

What I mean as Active/Active
Our original configuration
Issues we had running the application in Single Node.
Why we went to Active/Active over 10 years ago
Why does Active/Active work for this application and why it might not work in your environment
How we manage our Active/Active environment including distribution of external resources
How do we manage migration to single node
File transfer and replication
Major benefits of Active/Active
Issues we have with Active/Active
What the future holds.

BC-2956-U
Hot-Hot for Dummies
Marty Edelman, Creative System Software
Audience Level: Introductory

In today’s competitive environment disaster planning is no longer a luxury but a necessity. Minutes of downtime can cost millions of dollars and drive customers to competitors. Using industry-standard tools and building upon an SOA infrastructure, learn how one of the largest retailers in the world implemented a Hot-Hot infrastructure using a few simple library calls. Learn in depth how these libraries were built, how they handle real-time, near real-time, and offline transactions, how recovery is automated and how response time only increased a few milliseconds. The benefits of the system will be explained along with an ROI case study that can be used to help justify this system for your own application.