In the rush to design secure systems that respond to the business needs of the organization, the ability to recover systems in the event of failure might be overlooked. It should not be. Recovery processes, including emergency management of remote servers, are often seen as a network administration process and are not included as part of a security designer's purview. However, one aspect of security is ensuring the availability of data and systems, and dealing with the risk of systems failure—whether due to hardware or software malfunction or attack—is something that should be a part of every security design. Use the following guidelines to help you design for recovery:
Understand the capabilities of the systems for recovery. All systems should
be considered, and a comprehensive strategy should be developed for each. How?
ever, each platform presents specific benefits and challenges. Windows Server
2003 offers many features to assist in recovery, including volume shadow copy,
backup and restore processes, Automated Systems Recovery (ASR), and Emer?
gency Management Services.
Divide systems by computer role on the network. What a computer is used
for dictates what type of backup and restore plan needs to be developed for it.
Databases, for example, might contain extremely sensitive data, large amounts of
data, require special backup software, and take a long time to back up and
restore. Desktop computers, on the other hand, might not store any data locally
and might simply require a fresh imaging or replacement should they fail.
Have written procedures for each computer role. Don't trust any one
person's knowledge of how computers must be backed up or how they can be
restored. This person might not be available when needed. Written plans are
critical.
Practice recovery steps. Provide test systems, and require IT personnel who
recover systems to practice the recovery steps. Keep records of which IT person?
nel have completed practicing the recovery steps. Provide a drill. Simulate the
loss of a domain controller, for example, and require IT personnel to recover the
system. Provide feedback about how well they responded to the drill. If systems
are lost and then recovered or not recovered, provide postmortem discussions on
the process.
Determine when to use volume shadow copy. This service empowers the
end user with the ability to recover previous editions of files or to restore files that
have been accidentally deleted.
Design a backup strategy. Windows Server 2003 has backup and restore pro?
cesses built into it, and other companies' products are available as well. Determin?
ing which product to use will depend on the availability of the services necessary
to your design. Backup strategies include decisions to be made about frequency,
comprehensiveness, storage of backup media, offsite storage, personnel, logging,
number of copies, security for backup media, and audits of backup processes.
Rank systems to determine how critical they are. In a formal business continuity plan, a business impact analysis (BIA) is constructed to determine –which systems are the most critical to an organization's survival. The BIA is an interdepartmental project. Without this information, an IT design for recovery might consist of the identification of critical IT resources such as domain controllers, messaging servers, firewalls, and the like. The reason for this qualification is to create more comprehensive recovery plans for these systems and to determine the order of recovery efforts should multiple systems be affected at the same time.
Exam Tip Recovery of systems at remote locations can be accomplished via the use of Emergency Management Services (EMS) and specialized hardware. Providing such access to systems could result in malicious activity if the systems are not properly secured. EMS, auxiliary hardware connections, or both could be used to connect to and disrupt the normal services provided by the server. Any design that provides such access should include security.
Determine special systems strategies that are the result of the server role
or location. Domain controllers, e-mail servers, firewalls, Internet Information Services (IIS), Domain Name System (DNS) servers, and other network server roles require backup and recovery techniques in addition to the backup and recovery required by the base operating system. Servers located at remote locations might require processes as provided by the Emergency Management Services available with Windows Server 2003.
Complete the recovery process plan. All the information gathered is formulated into a comprehensive plan. The plan includes procedures and schedules for each type of product and systems—both backup and recovery. It also defines the order of recovery during times of multiple failure, assignment of responsibilities, and a plan for continuous testing and auditing of recovery plans.
Test each piece of the plan. Testing each piece of the recovery process
ensures adequate backups are being prepared and that the procedures are valid.
Testing might also show that alternative procedures or additional processes need
to be added.
Audit backup processes. Designing an adequate plan is a good start, but even
better are plans that are actually implemented, maintained, and reviewed. There
should be processes in place—including logging—to monitor backup.
Practice: Creating the Security Design Framework
In this practice, you will predict threats to a company, design a segmented network, and design a recovery process for a fictitious company. You will then match security plan elements to parts of a security framework and identify devices that segment internal networks into zones. If you are unable to answer a question, review the lesson materials and try the question again. You can find answers to the questions in the "Questions and Answers" section at the end of this chapter.