22.12.2012 Views

Front cover - IBM Redbooks

Front cover - IBM Redbooks

Front cover - IBM Redbooks

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

146 Lotus Security Handbook<br />

At this point we have defined the four types of security zones and the various<br />

types of firewall functions we can implement at the zone boundaries. The last<br />

stage of the model involves defining the criteria we will use to determine what<br />

zone a particular system should be located in, and what firewall functions we<br />

need to provide on the data flows between zones. We will be able to make these<br />

determinations by defining our data flow policies, meaning how must we protect<br />

data of various classifications that is traversing two zones.<br />

4.2.4 Interzone connectivity: Data flow policies<br />

Now that we have defined our four types of security zones and the firewall<br />

functions we can employ at the various boundaries between any two adjacent<br />

zones, we next need to know how to apply these building blocks in a practical<br />

network architecture design. We need to define rules for connectivity between<br />

zones that support our organization’s security policy.<br />

In fact, the data flow policies are an integral part of our security policy. We do not<br />

want to undermine or bypass the business purpose behind separation of<br />

resources into specific zones by providing improper connections. We need<br />

specific security controls for communication between different zones. Some of<br />

the controls we will require are provided by the zone boundary firewall devices,<br />

while other controls must be provided by the hosts, the host applications, or both.<br />

We first describe the criteria we use to categorize data flow policies. So what do<br />

we mean by data flow policies? We simply mean we have policies or rules for a<br />

client in “zone x” to communicate with a server in “zone y.” This is assuming a<br />

client in zone x is permitted to connect to a server in zone y under at least one<br />

set of criteria. Keep in mind that data flows are defined uni-directionally. The<br />

requirements for one direction of data flow may be different for flow in the<br />

opposite direction. Our data flow policies are directionally dependent.<br />

In our four-zone model, we allow for multiple Proxy zones and Data Access<br />

zones. There is by definition only one Internet zone, and for practical purposes,<br />

one Intranet zone. In some organizations, there are multiple Intranets, but in our<br />

experience this is usually not by intention but is a product of circumstances such<br />

as mergers or international organizational separations. Although the typical<br />

Intranet has several subnets and routers, there is normally little to no filtering<br />

being done within the Intranet itself. Figure 4-8 depicts the logical connections<br />

we are going to allow for our four types of zones.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!