Business Continuity/Disaster Recovery Planning

By Kenny Franklin, Professional Services Manager

How many of you have a comprehensive disaster recovery plan designed to minimize the effects on your business if a server fails?  If your building floods or burns down?  How about if your city experiences some sort of natural disaster? Think a disaster recovery plan is not important, or don’t have the time or resources to devote to developing one?   Recent events in New Orleans suggest being prepared is vitally important.

Simply having backup software/hardware and a rotation schedule for offsite storage of data is certainly one piece of the puzzle in that it addresses server recovery and data recovery.  It doesn’t necessarily answer other questions in the event of a disaster such as:

• Have you made decisions about who is responsible for what in the event of a disaster?
• Where/How will we procure new server hardware?  Can we get it quickly enough to meet our disaster recovery goals?  Do we need to keep spare equipment available at another location in order to meet recovery objectives?
• Where will my production activities re-locate in the event my building is not habitable?

Disturbing Statistics

As many as 80% of American companies have no DR plan

Somewhere around 90% of companies fail within one year of significant data loss

Roughly 40% of companies experiencing a disaster never reopen and roughly 30% close within two years.

Companies that are not able to resume operations within 10 days are not likely to survive

Developing a plan

Developing a plan involves determining corporate priorities. It is critical that top-level management participate in this process. Your company needs to develop disaster recovery targets not only for server/data recovery but also for revenue-generating activities to resume. For some businesses, getting revenue-generating activities running may take precedence over recovering data and computer systems. In other cases, revenue generation and computer systems may be more tightly linked. 

Another variable to consider is the cost of downtime. Figuring out how much it costs your company in lost revenue and/or lost worker productivity should be one of the first steps you should take toward developing a budget for your DR initiatives. How can you know how much to spend on disaster recovery process development and systems if you don’t know how much downtime costs your company? 

A plan should be in place for where your firm will take up operation in case your current building is uninhabitable. There are firms that provide backup office space and also provide some basic necessities such as telephones, desks, etc.

Technologies should be identified and tested that allow your firm to meet your recovery timelines. Below are several examples of backup/restore technologies and estimated recovery timelines:

• Simple tape backup – Once replacement hardware has been procured, 4-12 hours recovery time per server depending on the amount of data that needs to be restored. Tape restores are single-threaded processes, thus it will only be possible to restore one server at a time.
• Disk-to-Disk backup appliance with baremetal image capabilities – Once replacement hardware has been procured, 1-4 hours recovery time per server.  Many of these devices allow multiple servers to be restored simultaneously speeding overall recovery time.
• Boot from SAN and/or virtualized servers with data replication to offsite location – An hour or less for all servers once replacement hardware has been procured.

Documenting The Plan

Documenting your DR plan is critical. You don’t want to be relying on memory when in the middle of a disaster. This plan should become a living document reflecting changes as they occur both with internal IT systems as well as with critical business processes. All executive-level management and IT staff should have a copy of the plan. Things that should be documented include:

• Employee call list – a list of all employees home contact information to communicate
• Location of alternate work site
• A prioritized list of critical business processes and their associated recovery targets
• A prioritized list of servers, server applications,
• Recovery processes

Testing The Plan

The wrong time to find out there are flaws in your DR plan is when your company experiences an actual disaster. It is imperative that you test your plan regularly to make sure your targets are achievable and also to uncover potential flaws in the plan. And don’t forget to update your documentation as needed to reflect changes and new information.

Kenny Franklin is a Professional Services Manager at Polar Systems. He holds a degree in business from Cal State University in Hayward, specializing in management information systems. Kenny is also an accomplished jazz musician.