| UP | DOWN | ||
|---|---|---|---|
| Agilent Application Mix: Realistic Performance and Stress Resilience Testing in Modern Business, Triple-Play and Multi-Service Converged Networks | |||
| Kazakov,Phillip | Agilent Technologies Inc | ||
| Modern trends in networking resulted in many network devices becoming application-aware or performing conditional processing based on information contained in Layers 4-7 of the IP packets. Such devices are now inspecting the protocol headers as well as the data payload contained in application transactions: · Firewalls and intrusion detection devices parse the application data scanning it for attack or worm signatures · Layer 4-7 traffic management devices inspect the application headers and data payload to give priority to mission critical services and rate-limit the P2P or gaming traffic · Next-generation routers perform application-aware routing based on application and service types rather then Layer 3 destinations addresses. Such routers are capable of intelligent routing decisions like sending VoIP delay-sensitive traffic via low delay paths and delay-insensitive data traffic via higher delay paths. This in turn has changed the requirements towards the methodologies applicable to testing of application-aware network devices. Service providers and equipment manufacturers alike must now use real application traffic in protocol and transaction distributions representative of live networks to: · Measure the performance limits of network devices under real-world conditions with diverse network traffic and high volumes of users · Quantify the device scalability and plan network capacity to accommodate for expected traffic and user loads · Verify the stress resilience limits of network devices under peak traffic load or DoS attacks to ensure service availability and zero network downtime In this paper we will uncover the Agilent Application Mix test methodology that provides precise answers to these critical challenges. Application Mix uses a realistic mixture of data, voice and video application traffic complemented with DoS attacks and simulated malicious content such as worms and viruses. The traffic’s protocol distribution and application-layer transaction characteristics are based on extensive statistical analysis of traffic patterns measured in live networks. The methodology thus allows for subjecting devices to complex traffic conditions indistinguishable from real-world network environments, measuring the performance characteristics the devices will exhibit in real networks and assessing the peak levels of load and stress the devices will be able to withstand without disrupting service or compromising quality of user experience | |||
| Phillip Kazakov Product Manager, Electronic Measures Group, Data Networks Operation, Agilent Technologies. Phillip Kazakov is a Product Manager of Agilent’s Network Tester Layer 4-7 test system. Prior to joining Agilent, Phillip spent six years at Check Point Software, as an R&D Performance Lab Manager. Phillip also held the position of an IT Manager at the German Mission of IOM (International Organization for Migration), overseeing a corporate network with data, communication & security services. Phillip also spent four years with IBM Israel as a software engineer and later on as Project Manager, participating in different RS/6000, AS/400 and mainframe software development projects. | |||