Taking Complexity Out of Network Security – Simplifying Firewall Rules with TrustSec

Bruce Schneier, the security technologist and author famously said, “Complexity is the worst enemy of security.”

We have been working with some customers who agree strongly with this sentiment because they have been struggling with increasing complexity in their access control lists and firewall rules.

Typical indicators of operational complexity have been:

• The time that it can take for some organizations to update rules to allow access to new services or applications, because of the risks of misconfiguring rules. For some customers, the number of hours defining and actually configuring changes may be an issue, for other customers the biggest issue may be the number of days that it takes to work through change control processes before a new application is actually in production.
• The number of people who may need to be involved in rule changes when there are high volumes of trouble tickets requiring rule changes.

Virtualization tends to result in larger numbers of application servers being defined in rule sets. In addition, we are seeing that some customers need to define new policies to distinguish between BYOD and managed endpoint users as part of their data center access controls. At the same time, in many environments, it is rare to find that rules are efficiently removed because administrators find it difficult to ascertain that those rules are no longer required. The end result is that rule tables only increase in size.

TrustSec is a solution developed within Cisco, which describes assets and resources on the network by higher-layer business identifiers, which we refer to as Security Group Tags, instead of describing assets by IP addresses and subnets.

Those of us working at Cisco on our TrustSec technology have been looking at two particular aspects of how this technology may help remove complexity in security operations:

• Using logical groupings to define protected assets like servers in order to simplify rule bases and make them more manageable.
• Dynamically updating membership of these logical groups to avoid rule changes being required when assets move or new virtual workloads are provisioned.

While originally conceived as a method to provide role-based access control for user devices or accelerate access control list processing, the technology is proving of much broader benefit, not least for simplifying firewall rule sets.

For example, this is how we can use Security Group Tags to define access policies in our ASA platforms:

Being able to describe systems by their business role, instead of where they are on the network, means that servers as well as users can move around the network but still retain the same privileges.

In typical rule sets that we have analyzed, we discovered that we can reduce the size of rule tables by as much as 60-80% when we use Security Group Tags to describe protected assets. That alone may be helpful, but further simplification benefits arise from looking at the actual policies themselves and how platforms such as the Cisco Adaptive Security Appliance (ASA) can use these security groups.

• Security policies defined for the ASA can now be written in terms of application server roles, categories of BYOD endpoints, or the business roles of users, becoming much easier to understand.
• When virtual workloads are added to an existing security group, we may not need any rule changes to be applied to get access to those workloads.
• When workloads move, even if IP addresses change, the ASA will not require a rule change if the role is being determined by a Security Group Tag.
• Logs can now indicate the roles of the systems involved, to simplify analysis and troubleshooting.
• Decisions to apply additional security services like IPS or Cloud Web Security services to flows, can now be made based upon the security group tags.
• Rules written using group tags instead of IP addresses also may have much less scope for misconfiguration.

In terms of incident response and analysis, customers are also finding value in the ability to administratively change the Security Group Tag assigned to specific hosts, in order to invoke additional security analysis or processing in the network.

By removing the need for complex rule changes to be made when server moves take place or network changes occur, we are hoping that customers can save time and effort and more effectively meet their compliance goals.

For more information please refer to www.cisco.com/go/trustsec.

Follow @CiscoSecurity on Twitter for more security news and announcements.