By Mike Talon
In recent articles, we've discussed how to develop a disaster recovery plan that the powers-that-be will be willing to approve. After you get the green light for your DR budget, make sure you're ready to begin implementation.
The first problem you'll likely run into is one you can easily overcome—with the right preparation. Implementing a DR plan will almost definitely impact the servers in your production environment just as it will impact the DR systems themselves.
To help reduce this impact, you should plan for three significant issues ahead of time. Let's look at how you can properly prepare.
Deal with downtime
First, work around production schedules to find time for the inevitable outages. Outages are a mandatory part of almost every DR implementation. You'll need to install software, properly configure hardware, and test all components, which can mean a decent amount of downtime.
In most enterprises, admins can perform these tasks after-hours to lessen the impact. But it can be more difficult to find sufficient off-hour time in companies that operate 24/7, and users are more likely to feel the effects. In either case, make sure you properly alert employees about the outages, and don't forget to get prior management approval to back up these alerts.
While you can take steps to either minimize or eliminate the downtime, be sure you know what you're giving up in order to gain this minor advantage. Anything that works via integration with almost any file system will require a reboot or cause data to be unavailable for a short period of time.
Migrate your data
New hardware not only means outages—it also requires the migration of data (or the entire data system) to the newer devices. If you have to implement new hardware, make sure you properly plan for migrations.
Leveraging a combination of technologies could move the data to the new systems without too much service interruption. Replication and snapshot tools are famous for moving data from one system to another while it's still in use.
Backup tools can easily take data from one system and put it on another. However, they can typically move only older data, so they won't work in all situations.
No matter how you get the data from the first system to the next, you must find a way to reroute your end users to the new data system. This can be as easy as changing users' logon scripts and directories, or as complicated as changing code in software to find new servers. The ways to do this are myriad, so proper planning based on your systems is mandatory.
Educate your end users
After you properly plan for outages, you must notify end users exactly when and for how long the systems they use will be unavailable. Leveraging management approval can help limit the negative feedback you'll inevitably encounter when telling end users what to expect, but keep in mind that you're disrupting their workflow—and plan accordingly.
Find a way to communicate your efforts to end users to make sure they understand the new procedures you're putting into place. Users must know both what to expect during the implementation of the new systems and how to react to an emergency when the DR systems come into play. Proper dissemination of information on emergency procedures is also mandatory; users must know how to access your new systems once they go live.
Planning for implementation is a vital step toward properly executing your DR plan. Knowing how your business works and leveraging management intervention can go a long way toward properly implementing your DR systems.
Mike Talon is an IT consultant and freelance journalist who has worked for both traditional businesses and dot-com startups.