 |
 |
 Application Traffic Management
One of the most effective ways to get the most out of your IT investment is to optimize the application traffic moving across your network. A strategically sound approach to traffic management entails four closely related phases: classification, analysis, control, and reporting.
Traffic Classification
Today's top traffic management technologies offer application-intelligent classification capabilities that can identify hundreds of applications by their unique signatures automatically. Software like SAP, Citrix, Oracle, and KaZaA including applications that negotiate dynamic port assignments and masquerade to avoid detection can be identified, tracked, and sub-classified using various parameters. Among them:
- Advanced Layer 7 application signatures
- Sub-classifications including Oracle (by database), HTTP (by URL, MIME type, content type), Citrix (by published application or priority), and VoIP (by protocol or CODECs)
- Layer 3 IP addresses, address ranges, subnets, and host lists
- Layer 4 UDP and TCP ports, port ranges, and port lists
- QoS markings, including DiffServ, IP-ToS, IP-CoS, and IP precedence
- MPLS label and label ranges
- Frame relay interface, PVC/DLCI, ATM PVC, and ATM interface
- Ethernet MAC address, ISL-VLAN, 802.1q-VLAN, and 802.1p-LAN QoS Marking
Traffic Analysis
Traffic management software can provide detailed analysis of application performance and network efficiency describing peak and average bandwidth utilization, response times (divided into network and server delays), top users, top web pages, and top applications used.
Analysis of network behavior can include link-level intelligence that classifies by frame-relay or ATM PVC, IP subnet, MPLS paths, Ethernet MAC, and VLANs. Dozens of metrics can be tracked for each application, using response-time management (RTM) to permit advanced analysis of performance and network delays thereby enabling network and application groups to assign tasks appropriately and resolve problems more quickly.
Traffic Control
Policy-based traffic shaping, bandwidth allocation, and acceleration capabilities enable network administrators to take effective action based on the information they obtain.
Traffic shaping prioritizes critical application traffic, paces less critical applications, and contains unsanctioned and recreational applications. Minimum and maximum bandwidth levels can be specified, either per session or per application. And TCP and UDP rate control and sophisticated queuing capabilities provide the QoS tools needed to address performance issues:
- Proactively preventing congestion on both inbound and outbound flows
- Eliminating unnecessary packet discards and retransmissions due to queuing
- Actively addressing the end-to-end latency that plagues routers and other queuing devices
- Establishing a smooth, even flow rate to maximize throughput
- Extending policy control to packet marking (MPLS, 802.1 and ISL VLANs, DiffServ, and IP TOS/Precedecence), admission control (deny, discard, DoS containment, and redirection), and dynamic bandwidth and subscriber provisioning.
Application traffic can be accelerated and effective bandwidth and WAN capacity increased through intelligent compression of network traffic. What's more, application-intelligent compression maximizes those gains by applying the technologies that are most effective for specific applications. Latency management and TCP and UDP rate control technologies reduce end-to-end latency. Finally, costs for set-up and administration can be cut significantly by choosing solutions that simplify deployment and monitoring of the tunnel environment.
Traffic Reporting
Application traffic management is an ongoing process whose benefits will build as you continue to tune your network. This makes network utilization and application performance analysis critical to the success of your optimization efforts. Thorough application traffic reporting can also provide benefits related to service level agreements, enabling you to define performance standards, compare actual performance with service-level goals, and generate reports on SLA compliance.
|
 |
 |