Availability in a Converged World: Successful Providers Must Rise to New Market Standards
Originally published May 18, 2009
A new generation of telecom subscriber has emerged, and is here to stay. Savvy and educated, they demand choice, constant innovation, rich interfaces and new on-demand services wrapped in self-service packages. To keep these subscribers and reduce churn, providers need to be able to satisfy their demands for 24x7 services, available from anywhere, anytime, at broadband speeds. And if that were not enough of a challenge, subscribers are demanding location-based services and converged services that bridge voice and data networks.
A Telling ExampleAmp’d Mobile was a U.S.-based Mobile Virtual Network Operator (MVNO) that launched in 2005 with a highly successful marketing strategy aimed specifically at youth. The company saw significant growth, but its billing/OSS system could not scale to support it, resulting in highly publicized billing system failures. And in 2007, the company filed for Chapter 11 bankruptcy. The failure to scale, and other factors, had created massive subscriber churn in addition to billing and collection problems, ultimately bringing the company down. If Amp’ed were able to quickly scale the overloaded billing/OSS system to additional servers and maintain continuous availability, the company might have been able to save the day.
In introducing new services and supporting the convergence of existing products, TSPs are under constant pressure to upgrade and expand their billing and provisioning systems, exposing them to the challenges of frequent planned migration programs, resulting in yet another source of downtime. To today’s subscriber, downtime is downtime, whether it is due to system failure or routine maintenance, and today’s subscriber can easily switch providers if they so choose. According to Bain and Company, acquiring a new subscriber can cost 6-7 times more than retaining that subscriber’s business.
Continuous Operations in a Converged WorldIt’s a given that providers maintain continuous availability at the network level; but to compete in the current market, providers must be able to maintain similar levels of availability at the database and application levels. Therefore, it is critical that providers carefully evaluate data replication solutions because not all of them are able to support continuous availability during migrations, maintenance and upgrades or transfer data between heterogeneous, geographically dispersed sources. Providers must look beyond standard physical data replication or mirrored storage solutions to meet the recovery time and recovery point objectives of the databases that support critical billing and subscriber applications. Physical data replication cannot protect against data corruption issues and typically has distance limitations. And because they work in homogenous environments and the standby systems are not accessible, they fail to deliver the necessary cost-effective scalability to support the complex and disparate server and application environments required by next-generation TSPs.
True AvailabilityWhat TSPs require, then, is a non-intrusive software solution that can complement physical data replication to meet the availability requirements during planned and unplanned outages and enable true operational agility. Utilizing a heterogeneous and “transaction aware” data replication technology, for example, TSPs can capture each new committed transaction directly from a source database’s transaction logs with very low impact on the source system and deliver that changed data with sub-second latency across one or many heterogeneous systems – even in long-distance environments. This method allows the target system to be open and ready for immediate switchover and minimizes data loss – improving both the availability and recovery points for critical systems. In addition to this, TSPs are also able to minimize the risks involved in migrations – especially when migrating to quad-play services. By keeping the new system in synch with the old system while it is still actively used, the transactional replication solution limits the outage time to the seconds required for application switchover and allows continuous availability, even in complex and highly customized environments. With bi-directional data movement between the old and new systems, TSPs can have the option for immediate failback to the old system after switchover if the new system is not ready to support production users.
The same technology can address scalability issues that many TSPs face with growing subscribers and new service usage. Real-time heterogeneous data movement allows TSPs to offload some of the read-only activity, such as reporting, to a secondary, lower cost database that is continuously updated with the primary database. Some solutions can also offer bidirectional data replication and enable dual-active database implementations to support load balancing. Both of these implementations can free up the production systems for transaction processing and allow horizontal scaling cost effectively.
Thriving in a Competitive MarketIn an era when continuous operations and fast performance are not a luxury, but the norm, the market leaders will be the providers that optimize their infrastructures with real-time, transactional data replication capabilities – not just to meet current market needs, but to always be ready for what comes next.
Copyright 2004 — 2019. Powell Media, LLC. All rights reserved.
BeyeNETWORK™ is a trademark of Powell Media, LLC