Service Objects Weblog

Failover of Contact Validation Services, Ensuring the Continuous Flow of Your Data

January 24, 2012 at 2:12 pm serviceobjects2001 Leave a comment

Many companies talk about up-time and service level agreements that look great on paper but don’t perform up to customer expectations in a crisis. A good backup process, covering application software, is necessary and should be standard, but it’s not enough.

You need an architecture that provides failover from a primary server, to a back-up server. It should pick-up and provides the same service, with minimal interruption in the flow of data to your application. This is more complicated, and valuable, than a mere backup of the application or even load balancing. Proper failover can minimize the interruption of access to your contact validation services. For mission critical and business critical applications this involves automatic failover to a fully separate alternate location. This dual datacenter level of failover protects against single datacenter failure of servers, LAN and WAN network access failure and physical location failure modes such as fire, power, or natural disruptions.

When considering real-time contact validation services for a business application, where continuous uptime is critical, here are some things you should look out for:

• Live backup servers that perfectly mirror the live primary servers, they should be identical both in count and content
• Multiple datacenters, one hosting primary servers and the other backup servers, located in different locations of the country, if a disaster takes out one data center, or the network access, it should not affect the other
• 99.95% uptime, with assurances that both datacenters have never gone down at the same time
• With the application and content exactly the same in both production servers, doing a failover should be as simple as changing the URL from one server to the next. For example, you would code your application to automatically failover from the URL of the primary server to the URL of the back-up server based on a condition such as response timeout
• In the event of massive datacenter failure, your contact validation provider should redirect traffic to the backup server at the backup datacenter
• XML code should contain failover suggestions and any support team should be able to help their clients implement failovers

Having a data response failure in your live application, that utilizes contact validation, can mean unexpected losses as your customers usage of that applications disrupted. If a company is accustomed to having validated contacts imported into their CRM system or used in some application from their website, a loss of service, even temporary, can lead to corrupted or non-corrected data. Failure of your contact validation processes can cause sales people to waste effort following bogus leads or packages to be delivered to bad addresses, all of which could be avoided if the contact validation provider offers a business critical level of service as noted above.

Entry filed under: General Service Objects Information. Tags: back-up, backup, Contact Validation, failover, Service Objects.