Riverbed AppResponse is a modular platform. The foundational component is the Shark Packet Analysis (SPA) module. This base module provides the core functionalities of packet capture, storage, filtering, and deep packet inspection for network forensics and troubleshooting. All other specialized analysis modules, such as Application Stream Analysis (ASA) or Web Transaction Analysis (WTA), are licensed as add-ons that build upon the capabilities provided by the base SPA module.

B. UCA: Unified Communications Analysis (UCA) is a specialized, add-on module for monitoring and troubleshooting voice and video performance, not the base module.

C. WTA: Web Transaction Analysis (WTA) is an add-on module that provides deep monitoring of end-user experience for web and SaaS applications.

D. ASA: Application Stream Analysis (ASA) is a powerful add-on module for real-time application performance monitoring, but it requires the base SPA module to function.

1. Riverbed AppResponse 11 Product Brief: "AppResponse 11 is a modular system. The base module is Shark Packet Analysis (SPA), which provides network forensics and troubleshooting. Add-on modules include: Application Stream Analysis (ASA)... Web Transaction Analysis (WTA)... Unified Communications Analysis (UCA)..." (Page 2, "Modular and Scalable" section).

2. Riverbed AppResponse 11.12 User's Guide: "The AppResponse 11 system is modular. The base module is Shark Packet Analysis (SPA). Add-on modules include Application Stream Analysis (ASA), Web Transaction Analysis (WTA), and Unified Communications Analysis (UCA)." (Chapter 1, "Introduction to AppResponse 11," Section: "AppResponse 11 Modules").

In Riverbed SteelCentral Portal Application Models, an "Internal Element" is the most granular component representing a specific, monitored application process or instance running on a host. These elements are the fundamental building blocks of an application's architecture within the model. For example, a specific Apache web server process, a Java Virtual Machine (JVM), or a database instance on a server would each be defined as an Internal Element. The model then maps the communication paths and dependencies between these specific instances to provide a detailed view of application performance and behavior.

A. Traffic between hosts is represented as a "Link" or "Flow" in the application model, not an element itself.

B. Servers are the hosts on which elements run; an Internal Element is the specific application instance on the server, not the server itself.

D. Logical sets of hosts are defined as "Tiers" (e.g., Web Tier, Database Tier), which are containers for multiple Internal Elements.

1. Riverbed SteelCentral Portal User's Guide, Version 4.8, Chapter 6: "Application models," Section: "Application model components," Page 127. The guide states: "Internal elements are the application instances that are part of the application model. For example, you can add a specific HTTP server as an internal element." This directly supports that an Internal Element is an application instance.

2. Riverbed SteelCentral Portal User's Guide, Version 4.8, Chapter 6: "Application models," Section: "Application model components," Page 127. The guide defines a Tier as: "Tiers are logical groupings of servers in an application model." This confirms that a logical set of hosts (Option D) is a Tier, not an Internal Element.

To calculate a utilization percentage, two values are necessary: the measured usage (numerator) and the total capacity (denominator). Riverbed AppResponse measures the actual traffic volume for a defined Host Group from packet data. However, it has no inherent knowledge of the total bandwidth capacity of the physical or logical link that the Host Group's traffic traverses. Therefore, the administrator must manually configure the Inbound and Outbound bandwidth values within the Host Group's settings. AppResponse then uses these user-defined capacity values as the denominator to calculate and report the Link Utilization Percentage metric for that specific group.

A. SNMP Polling is enabled: SNMP is used to poll metrics from network infrastructure devices (e.g., routers, switches), not to define the capacity for a logical Host Group within AppResponse.

C. WMI polling is enabled: WMI is used to collect performance metrics directly from Windows servers (e.g., CPU, memory), which is unrelated to the network link's bandwidth capacity.

D. Flow collection is enabled: While flow data provides information about traffic volume, it does not contain the total bandwidth capacity of the link, which is essential for calculating a utilization percentage.

1. Riverbed AppResponse 11.13 User's Guide, Chapter 4: "Configuring AppResponse," Section: "Defining Host Groups," Page 101.

The guide details the configuration options for Host Groups. For the "Inbound Link Bandwidth" and "Outbound Link Bandwidth" settings, the description explicitly states: "Used to calculate Link Utilization % for this host group." This directly confirms that specifying the bandwidth is the prerequisite for this calculation.

The Riverbed Portal prioritizes data sources to provide the most accurate representation of the end-user's experience. AppInternals JavaScript provides Real User Monitoring (RUM) data by capturing metrics directly from the end-user's browser. This is the most direct and highest-fidelity source for metrics like page load times and browser rendering performance. Therefore, when data from the AppInternals JavaScript agent is available for an application model, Portal gives it the highest priority for populating the End-User Experience Dashboard over network-inferred sources.

A. AppResponse ASA: Application Stream Analysis (ASA) provides valuable network and server delay metrics but infers performance from network packets, making it less direct than browser-based RUM.

B. NetIM SNMP Polling: NetIM provides infrastructure health metrics (e.g., CPU, memory) via SNMP. This data offers context but does not directly measure the application experience from the user's perspective.

D. AppResponse WTA: Web Transaction Analysis (WTA) reconstructs user web sessions from network traffic. While an excellent source for EUE, it is still network-based and considered secondary to direct measurement from the user's browser.

1. Riverbed Portal User's Guide: In the section describing Application Models and data sources, the documentation specifies the order of precedence for populating metrics. It states that for end-user experience metrics, data from AppInternals (specifically the JavaScript agent for RUM) is prioritized over data from AppResponse WTA, which is in turn prioritized over AppResponse ASA.

Reference: Riverbed Portal 4.7 User's Guide, "Application Models" chapter, section on "Data source precedence for application models," page 45. The guide explicitly lists the priority as: 1. AppInternals, 2. AppResponse WTA, 3. AppResponse ASA.

2. Riverbed NPM and APM Solution Guide: This official vendor documentation outlines the roles of different products. It describes AppInternals as the solution for deep application diagnostics and real user monitoring, positioning its data as the primary source for user-centric performance analysis within the integrated suite.

Reference: Riverbed Application Performance Management (APM) Solution Brief, section on "End-user experience monitoring," which details the function of the JavaScript agent in capturing client-side metrics.

AppResponse functions by capturing and analyzing network packets to derive performance metrics. It calculates metrics such as Round Trip Time (RTT), TCP Retransmissions, and Server Delay directly from this packet-level analysis. This rich performance data is then exported as Application Flow (AppFlow) records to NetProfiler.

Conversely, Network Device Interface Utilization is a metric typically obtained by polling network infrastructure like routers and switches directly via the Simple Network Management Protocol (SNMP). While NetProfiler can be configured to perform this SNMP polling to gather utilization data, this metric is not generated by AppResponse's packet analysis nor is it forwarded by AppResponse to NetProfiler.

B. Round Trip Time: This is a core network delay metric that AppResponse calculates by measuring the time between TCP SYN and SYN/ACK packets and is a key component of the data it forwards.

C. Retransmissions: AppResponse identifies TCP retransmissions by analyzing packet sequence numbers, providing crucial insight into packet loss and network health, which it reports to NetProfiler.

D. Server Delay: This is a critical application performance metric representing the time a server takes to process a request. AppResponse measures this from packet timings and includes it in the flow data sent to NetProfiler.

1. Riverbed SteelCentral NetProfiler User's Guide: In the section describing data sources, the guide distinguishes between flow data from AppResponse and SNMP data. It states, "AppResponse appliances provide the richest flow data, called Application Flow... Application Flow includes... performance metrics, such as... Network time [RTT], Server time... Retransmission counts." (Reference: SteelCentral NetProfiler 10.9 User's Guide, Chapter 2: "Getting Started," section on "Data Sources").

2. Riverbed SteelCentral AppResponse User's Guide: The documentation details the metrics available through packet analysis. It explains that "AppResponse breaks down end-user response time into contributing sources of delay, including server delay and network delay (round-trip time)." It also describes the monitoring of TCP health metrics like retransmissions. (Reference: SteelCentral AppResponse 11.8 User's Guide, Chapter 1: "Introduction," section on "Key Features").

3. Riverbed Command-Line Interface Reference Manual: The manual's sections on configuring flow export from AppResponse (flow-export) list the specific metrics derived from packets that are included in the AppFlow records. These lists consistently include RTT, server delay, and retransmission-related metrics, but do not include device-specific SNMP metrics like interface utilization. (Reference: Riverbed SteelCentral AppResponse Command-Line Interface Reference Manual, section on flow-export commands).