2 cissp ® Official Study Guide Eighth Edition

Service Bureaus
A 
service bureau
is a company that leases computer time. Service bureaus own large server 
farms and often fields of workstations. Any organization can purchase a contract from a 
service bureau to consume some portion of their processing capacity. Access can be on site 
or remote.
A service bureau can usually provide support for all your IT needs in the event of a 
disaster—even desktops for workers to use. Your contract with a service bureau will often 
include testing and backups as well as response time and availability. However, service 
bureaus regularly oversell their actual capacity by gambling that not all their contracts will 
be exercised at the same time. Therefore, potential exists for resource contention in the 
wake of a major disaster. If your company operates in an industry-dense locale, this could 
be an important issue. You may need to select both a local and a distant service bureau to 
be sure to gain access to processing facilities during a real disaster.
Cloud Computing
Many organizations now turn to cloud computing as their preferred disaster recovery 
option. Infrastructure as a service (IaaS) providers, such as Amazon Web Services, 
Microsoft Azure, and Google Compute Cloud, offer on-demand service at low cost. 
Companies wishing to maintain their own datacenters may choose to use these IaaS options 
as backup service providers. Storing ready-to-run images in cloud providers is often quite 
cost effective and allows the organization to avoid incurring most of the operating cost 
until the cloud site activates in a disaster.

---

Recovery Strategy 
825
Mutual Assistance Agreements
Mutual assistance agreements
(MAAs), also called 
reciprocal agreements
, are popular in 
disaster recovery literature but are rarely implemented in real-world practice. In theory, 
they provide an excellent alternate processing option. Under an MAA, two organizations 
pledge to assist each other in the event of a disaster by sharing computing facilities or other 
technological resources. They appear to be extremely cost effective at first glance—it’s not 
necessary for either organization to maintain expensive alternate processing sites (such as 
the hot sites, warm sites, cold sites, and mobile processing sites described in the previous sec-
tions). Indeed, many MAAs are structured to provide one of the levels of service described. 
In the case of a cold site, each organization may simply maintain some open space in their 
processing facilities for the other organization to use in the event of a disaster. In the case of 
a hot site, the organizations may host fully redundant servers for each other.
However, many drawbacks inherent to MAAs prevent their widespread use:
■
MAAs are difficult to enforce. The parties might trust each other to provide support 
in the event of a disaster. However, when push comes to shove, the nonvictim might 
renege on the agreement. A victim may have legal remedies available, but this doesn’t 
help the immediate disaster recovery effort.
■
Cooperating organizations should be located in relatively close proximity to each other 
to facilitate transportation of employees between sites. However, proximity means that 
both organizations may be vulnerable to the same threats. An MAA won’t do you any 
good if an earthquake levels your city and destroys processing sites for 
both