Business dependency on IT systems has never been greater, so when things go wrong the impact can be catastrophic. Messaging systems today, for instance, are a critical component in business activity. Whether it’s the CEO waiting for information on her BlackBerry as part of an acquisition negotiation, or the thousands of queries that flow through the corporate website and are dealt with by e-mail, any glitch in the messaging infrastructure can wreak havoc on productivity, revenue and the organization’s reputation.
Yet despite the critical role of business applications, the typical approach to keeping them up and running remains in the dark ages for many companies. Mostly, corporate e-mail systems are protected through backup, resulting in extended recovery times should things go wrong. And, of course, messaging systems don’t exist in isolation. What seems like a simple e-mail system actually touches many applications in the typical deployment. Alongside e-mail servers such as Microsoft Exchange or Lotus Domino, there will be BlackBerry Enterprise Servers, SQL configuration databases and content sources such as file systems and collaboration portals. So, even if the e-mail server is up and running, this doesn’t mean that the messaging system is able to do its job.
In today’s competitive and budget-sensitive world, nothing short of continuous availability is required for messaging and related systems.
What we need is an approach that delivers continuous availability and that is designed to meet 21st century challenges. Areas to consider include:
1. Is high availability alone sufficient? How would business be impacted if a site-wide issue took the messaging system down? Without a disaster recovery site capable of delivering a near-zero recovery time objective (RTO), the impact on the business of a power-outage would be significant. Consider looking at systems that can deliver a combination of local high availability and remote disaster recovery in an integrated fashion.
2. Are bandwidth costs a barrier to disaster recovery? Maintaining a continuous clone of the e-mail server off-site, using host-based replication and automated failover, is a good option for many organizations. But bandwidth consumption in some such solutions could be an issue. Look for ways to decrease bandwidth traffic, perhaps with on-the-fly de-duplication, as a way of overcoming cost barriers. If the de-duplication is built into the disaster recovery solution rather than requiring extra hardware, then so much the better.
3. Can you rely on staff always being available? Some organizations look to do what they think is the right thing and use pure data replication or log-shipping as a protection method. However, this relies on users reporting issues, IT investigating the problems, and the recovery of e-mail servers and reconnection with the database, all of which can take considerable time. Consider the scenario of a CEO in London on a Monday morning needing immediate access to e-mail, yet it is Labor Day in the U.S. and also midnight in California (where the problem is located). It’s substantially better to automate as much as possible, keeping the CEO working, than spending small hours rebuilding systems.
4. Is immediate virtualization the most cost-effective approach? In many companies the e-mail servers are stable and running on fully depreciated hardware. Adopting virtualization for messaging continuity will likely demand the introduction of shared storage, new servers to be bought and disruption during migration. Look at whether you can extend the life of the physical servers by using virtualization as a business continuity platform serving multiple e-mail servers. Think of it as server consolidation for the high availability servers. This can prove a great way of testing compatibility between physical deployments and future virtualization.
5. Mobile messaging is not all about BlackBerry. Yes the BlackBerry Enterprise Server is an important component in e-mail flow. But if the e-mail server is down, or SQL configuration databases have crashed, then the smart phone itself is no more than an expensive gadget.
In addition to these five points, consider this: a recent study by Osterman Research showed that in organizations of 1,500 people, the annual productivity loss caused by e-mail downtime is estimated to be over $250,000 per year. In this economy, that statistic is nothing to scoff at.
Waiting for an e-mail outage to disrupt, or even destroy, your organization is not an option. The time to address business and messaging continuity -- and come out of the dark ages into the 21st century -- is now.
Andrew Barnes is Senior Vice President of Corporate Development for Neverfail.
