Healthy VoIP Nets�Part XXVI�Network Management Architectures: Apparent Networks
Apparent uses a unique set of techniques to assess the condition of network paths from users to applications.
Apparent Networks, headquartered in Wellesley Hills, Mass., develops and markets software designed to give enterprises and managed service providers a radically different and more effective way to assess the performance of their converged networks.
Founded in 2000, Apparent Networks is a privately held firm backed by top-tier venture investors, including Bain Capital Ventures and JMI Equities.
The Companys software gives network management teams and their line-of-business constituents the ability to track the performance and service quality that their networks are deliveringnot just network availability.
Apparent Networks presently employs about 50 people, maintains a large development center in Vancouver, B.C., and has several sales and service locations throughout North America.
Its software solutions have been successfully deployed by large enterprises, managed service providers, and tier-one technology vendors including 3Com, Boeing, Electronic Arts, Fujitsu, Hewlett-Packard, IBM, Network Appliance, Nortel, Polycom, Siemens, Symantec, and Toshiba.
Apparent Networks claims that its flagship software offering, AppCritical, provides a radically different basis for assessing, measuring, and diagnosing converged networks. Unlike traditional network management tools that look in detail at individual devices, AppCritical utilizes a patent-protected expert technology for network behavior, providing network diagnostic, assessment, measurement, and monitoring functionality that the company feels is markedly superior to that provided by the traditional network management products.
This system gives users complete visibility into their network infrastructures, and analyzes the entire end-to-end network path used by an application without any breaks or blind spots. This includes the ability to see inside the service providers networks to determine exactly how that network is performing and whether or not it is affecting the application traffic and end users.
The AppCritical product includes several distinguishing features.
First, it provides Real-time, End-to-end Monitoring, continuously monitoring the capacity and quality of thousands of network paths in real time, including extending this monitoring and visibility through service providers networks to the end users.
Second, the system provides Active Path Assessment, which automatically transmits small, precisely timed packets through network paths using the Internet Control Message Protocol (ICMP) or the User Datagram Protocol (UDP). As these small, synthetic packets traverse the network, their timing is distorted by delays, restrictions, and limitations, all of which yield distinctive markers carried back by the packets.
AppCritical analyzes these markers, creating a virtual image of the present state of network paths, and determines whether or not they meet requirements of network-dependent applications.
Key network path measurements that the system captures and analyzes include: host name and address; measured and reported Maximum Transmission Unit (MTU) size; achievable bandwidth; best, average, and worst Round Trip Time (RTT); propagation delay; packet jitter, packet loss, and packet reordering; Mean Opinion Score (MOS); and utilization.
Third, its Diagnostics use patented algorithms and expert-system analytics to pinpoint the exact cause and location of any problems. Providing a path-centric perspective, the system determines whether the problem is network-related or application-related, and if its a network problem, exactly what is causing it, where it is occurring, and what needs to be done to resolve it.
The primary network faults detected by AppCritical include: congestion, media errors, collision domain violations, compromised Quality of Service (QoS), rate limiting queues, MTU conflicts, and black hole and grey hole hops.
The AppCritical architecture consists of three main components.
Sequencers are small pieces of code that can run on Windows, UNIX, or Linux environments, and are placed at strategic network locations (like data centers), where they create test packets and use them to conduct network path tests and collect real-time network performance data.
The Network Intelligence System manages the Sequencers and controls all path testing; analyzes data collected from Sequencers with identification, correlation, and remediation analysis engines; and presents the results to the Operations Console, which stores all data and analyses in a database.
Finally, that Operations Console presents the consolidated information about path and network application performance in simple, intuitive and actionable formats, using common language.
AppCritical also automatically generates detailed reports, including charts, graphs, and summary views. These customizable reports may include a test summary (see Figure 1), historical path service quality charts (see Figure 2), and a current snapshot of current network conditions (see Figure 3). These reports establish a common ground from which teams can resolve pressing problems and collaborate more effectively on strategic initiatives.
Further details on the Apparent Networks architecture and products can be found at http://www.apparentnetworks.com/. Our next tutorial will continue our examination of vendors network management architectures.
Copyright Acknowledgement: © 2008 DigiNet Corporation®, All Rights Reserved
Mark A. Miller, P.E. is President of DigiNet Corporation®, a Denver-based consulting engineering firm. He is the author of many books on networking technologies, including Voice over IP Technologies, and Internet Technologies Handbook, both published by John Wiley & Sons.