June 8, 2015
“Disaster” is a word that no business on this planet likes to hear, especially when the word refers to events that directly result in losses of assets and revenue. Both tangible and intangible assets — like data and physical infrastructure – are the lifeblood of enterprises. If something were to happen to the IT infrastructure of even the smallest organization, you can expect huge delays and operational issues, potentially leading to large revenue losses.
Despite all of this, it appears that disaster recovery, failover implementation and risk mitigation are often considered “fat cats” that require significant budget allocations. Smaller businesses are much less likely to adopt these solutions because they see them as too expensive and often unnecessary.
However, cloud-based technologies have allowed even the smallest of organizations to create scalable infrastructures that empower them with a chance to interact with the same platforms that bigger businesses have been using for years. These include enterprise resource planning solutions, customer relationship management software and other applications for collaboration, communication and automation. With the various cloud delivery models, such solutions have never been more accessible. But with the increasing level of reliance on such business platforms, now is the time to start seriously considering disaster recovery mechanisms.
What Is A “Disaster?”
In the context of disaster recovery, a disaster is any event that occurs either in the physical or digital realm that result in losses in IT systems, and can interrupt the ability to conduct operations. These can be a result of natural calamities that can damage physical infrastructure, or man-made concerns, such as accidents, civil unrest, fire, or power failures.
In terms of IT, these can involve switch failures, server shutoffs, hard drive crashes, and ISP disconnections. All of these things happen to businesses every day. InformationWeek reports that IT downtime can cost as much as $26 billion annually. In addition to this, the report estimates that 56 percent of enterprises in North America have poor disaster recovery plans. There are also additional costs that add up in the long-term, particularly with regard to lost reputation that often comes with perceived instability.
Disaster Recovery Is More Important Than Ever
Fifteen years ago, you could easily get away with server downtime and unreliable connections. Back then, it wasn’t the end of the world if your IT systems cease to operate due to downtimes. Today, however, a greater proportion of business operations have digitized their business processes. Thus, connectivity will mean more to continued success than it has in the past. Taking care of your online presence should now be priority No. 1.
If you depend on online sales, there’s absolutely no reason why you shouldn’t have an appropriate disaster recovery plan, period. Depending on your industry, you may also need it to conform to regulations such as the Sarbanes-Oxley Act, the IRS Procedure 86-19, the Consumer Credit Protection Act, HIPAA, or ISO 9000.
What Are Your Options?
In terms of ensuring the integrity of data, there are multiple disaster recovery techniques that can enable a business to respond to any disaster scenario. While there are inevitably incidents that are out of reach and out of control, effective disaster recovery mechanisms will involve factors that are within the control of an organization. Preventive measures that can mitigate or prevent disasters are especially important. This minimizes your chances of “catastrophic failure” which, in laymen’s terms, means a complete inability to provide services to your customers online.
A disaster recovery plan should include a plan of action that responds appropriately to loss or damage of mission-critical IT systems. These include the likes of load balancing (a preventive measure), as well as reliable replication methods that copy data over to a distributed server infrastructure, to ensure that data is safe even if one of these servers fail for some reason.
There are three methods in replicating data:
- Synchronous replication is the most secure and the most reliable, albeit the most resource-hungry method. Data is copied byte-by-byte and the process continues until the destination server confirms that it has both received and written the data.
- Asynchronous replication should be used only in situations where a bit of data loss is acceptable. With this method, the source server sends data indiscriminately to its destination. It doesn’t wait for a confirmation that the data has been received, let alone written. The advantage of this method is that you save resources and bandwidth. It is faster, albeit less reliable.
- Semi-synchronous replication is the compromise between synchronous and asynchronous replication. The source server will copy data and it will wait for a confirmation of reception from the destination. It won’t wait for the destination server to write the data to disk, though. The data will eventually be written, but it is temporarily stored in a memory buffer if the write process gets overwhelming. The data can be, for example, queued in RAM to be written in a few seconds while it deals with more pressing matters.
The good news is that disaster recovery and failover mechanisms do not need to make a big dent in the budget. An eCommerce website, for example, can enjoy the benefits of load balancing, Web application firewall and content delivery network for a monthly subscription that does not exceed $60.
Doing business online isn’t as simple as keeping a website up anymore. Your customer database has to be connected on all fronts, and you need to keep everything running smoothly. To do this, you need a way to get things back up and running within an acceptable amount of time – your “recovery time objective” – should disaster strike. There’s nothing worse than that feeling of powerlessness when your systems won’t work anymore and you have no defibrillator to revive your business. Don’t let yourself get in that situation. It’s better late than never to take the first steps toward becoming a disaster recovery ninja.