Create a Location-Based Traffic Tree with Global Applications

Instructions for organizing the traffic on your network by location, but applying policies by application

A location-based traffic tree with global applications is appropriate for a main site's WAN or Internet link with traffic that goes to multiple branches. It categorizes traffic first by travel direction, then by location. Then it applies per-application policies to each location's application traffic, borrowed from root-level inheritable classes listed below the location classes. But all traffic from one application is managed with one strategy, independent of branch. The metrics, graphs, and reports are the same as those available in the simple location-based tree. This tree is most common when branches have similar business priorities because each application is managed the same way for all locations. This traffic tree is the most difficult to understand conceptually, but the easiest to configure and needs the least amount of traffic classes, matching rules, and configuration time. It provides a very scalable and easy to manage main-site tree for traffic shaping but does limit visibility. This type of tree is not appropriate for PacketShapers without the control module, as the only purpose of the application classes at the bottom is to supply policies. For more information on this type of tree's capabilities, limitations, scaling considerations, and configuration recommendations, see Traffic Tree Designs.

The following steps help you create a location-based traffic tree with separate locations and applications.

Steps:

1. Enable traffic discovery, telling PacketWise to automatically create classes for applications and protocols as traffic passes.
   
   Let some time elapse to give PacketWise a chance to discover the various applications that travel your network over time.
   
   
2. Disable traffic discovery, telling PacketWise to stop classifying traffic.
   
   
3. Remove the traffic classes for types of traffic you do not want to manage separately and for which it's okay to simply count the traffic in with a default class.
   
   Note: The previous two steps are optional. You can leave traffic discovery on and refrain from pruning your classes, if you wish.
   
   
4. If you want to refine application traffic classes, consult Classification Hints and Examples for ideas.
   
   
5. Create a new class for each of your locations under both the Inbound and Outbound branches of your tree. When you create your matching rules, use criteria to identify your locations (usually subnets, host lists, addresses, or VLAN or PVC identifiers). (Example)
   
   You can create each class manually. Or, if there are many classes and it gets to be a tedious task, make it easier by creating a PacketWise command file and a data entry form.
   
   PacketWise automatically places your location classes above your application classes in the tree. There is no need to make any exception classes.
   

Note: Although your traffic tree is now complete, remember that additional steps are needed to implement your control strategy for this type of tree. The PacketGuide recommendation Control Branch Traffic from a Main Site has details.

PacketGuide recommendations that describe creation of a specific type of traffic tree:

• Create an application-based traffic tree
• Create a simple location-based traffic tee
• Create a location-based traffic tree with per-location applications
• Create a location-based tree with global applications