Let There Be Light! Using HDX Insight to Gain Visibility into XenDesktop and XenApp

HDX Insight is an invaluable tool that provides organizations using NetScaler to deliver XenDesktop hosted virtual desktops and applications with unprecedented visibility into the ICA traffic on their networks. Part of NetScaler Insight Center and complementing Web Insight, HDX Insight collects, correlates and displays highly informative network usage and performance data for XenDesktop/XenApp, both in aggregate and down to the level of individual user sessions. This data helps both network and application administrators to identify and triage network and application issues instantly.

The Beauty of HDX Insight

Key features and capabilities that enable HDX Insight to deliver unparalleled value include the following:

  •  Simple, cost-effective deployment. Unlike other solutions which have multiple touch points and require a complex configuration of network taps or intrusive agent software, HDX Insight provides end-to-end visibility with little more than an organization’s existing footprint of NetScaler ADCs, NetScaler Gateways, and CloudBridge WAN optimizers. The required instrumentation (AppFlow) is already a part of those solutions; it only needs to be turned on and pointed to an instance of the NetScaler Insight Center virtual machine (VM). Furthermore, the NetScaler Insight Center VM is a self-contained solution that incorporates all necessary sub-components, including an integral database.

Figure 1: High-level Solution Architecture
(source: white paper)

  •  Output = actionable information. Instead of a pile of reports featuring raw performance data, with HDX Insight IT teams obtain granular and actionable information based on the automated correlation of network and application level data. Everything needed to pinpoint the source of performance problems is provided in one convenient place, avoiding the need to manually piece a picture together by flipping between multiple consoles.
  • Granular ICA details. Unlike solutions that only recognize the presence of ICA traffic, HDX Insight not only decrypts but also parses ICA sessions. The result is a unique degree of insight down to the level of individual virtual channels.
  • Combination of real-time and historical data. HDX Insight avoids the need to resort to multiple solutions by addressing enterprise needs for both real-time and historical data in a single package.
  • Seamless integration with Director 7.0 and above. All HDX Insight data and reports can be viewed from within the same console organizations use to manage their XenDesktop environments.

What really sets HDX Insight apart from conventional solutions, however, is: (a) the breadth and depth of the information it supplies, and (b) what IT teams can do with that information.

Breadth and Depth of Available Information

High-level aggregate information with drill-downs to session-level data can be accessed from several different perspectives, including:

  • Users. Administrators can view a snapshot of the average round trip time (RTT, or screen lag time), WAN latency, datacenter latency, and bandwidth consumption across all users over select time periods (e.g., last 5 minutes, hour, day, week or month). The same screen can also be used to view Top Users by a range of attributes (e.g., number of apps/desktops launched, bandwidth consumed, latency and number of active desktops/apps) or search for individual users.
  • Applications. In addition to the aggregate bandwidth being consumed by all hosted applications, this view provides an app-by-app breakdown that includes the number of active sessions and apps, total count of sessions and apps launched, and launch duration across select time periods.
  • Desktops. In addition to total bandwidth being consumed by all hosted desktops, admins get a breakdown by user that includes the number of active sessions and desktops, session duration, bandwidth utilization, RTT, datacenter latency, and WAN latency.
  • Gateways. The aggregate/summary data for this view is based on the traffic transiting each individual instance of NetScaler/NetScaler Gateway (or CloudBridge) and provides a way to compare performance across sites/locations.

As discussed above, HDX Insight is also the only solution that reliably dissects the ICA protocol to provide granular details of what’s going on at each virtual channel level.

 

Figure 2: High-Level Summary for Desktop Information

Of course, the true measure of HDX Insight is what it enables IT teams to do. So let’s take a look at a handful of common use cases.

#1a: Tactical Troubleshooting – RajivS Reports Slow App(s)/Desktop

The scenario in this case is an individual user, RajivS, reporting a performance problem. The entry point for the investigation is the summary screen for User information.

In addition to the aggregate user information, here we can also find the summary level information for individual users, including RajivS. What we notice right away is a relatively high RTT (screen lag time) 819ms. We can also see that the WAN (12ms) and datacenter (1ms) latencies are low, suggesting the source of the problem lies elsewhere. In fact, by drilling into the record for RajivS and hovering over the RTT for one of the active sessions for RajivS, we can see that the host delay is a whopping 587ms. So the problem is clearly with the server in this case, the IP address for which we can readily see in the session record as well.

 

#1b: Tactical Troubleshooting – markt Reports Slow App(s)/Desktop

In this second example, we have a similar situation, this time with “markt” reporting a slow experience. In this case, however, we immediately see from the summary screen for User information that the client network is the main culprit.

 

 Drilling into the summary level record for markt further reveals high WAN latency for all of markt’s sessions. As a result, the support desk should coach “markt” on options for finding a higher performing network service/connection.

 

 If necessary, the network admin/support desk can also take a look at the individual applications markt is accessing during his current session.

 

#2 Proactive Performance Management

Investigations similar to the ones described above can also be performed to proactively keep developing issues from becoming bigger problems. For example, an administrator that notices an entire group of users with high RTTs all stemming from delays with the same host could troubleshoot and remedy the host-level problem in advance of receiving calls from the affected users.

In a similar manner, routine review of the information provided by HDX Insight might reveal a trend that requires proactive attention – such as adjusting network QoS settings in response to steadily increasing volumes of video traffic.

 

#3 Capacity/Strategic Planning

HDX Insight can also be used for capacity planning of network and application infrastructure. For example, data pertaining to concurrent apps, desktops and sessions and related bandwidth consumption can be used to plan/size upgrades to LAN/WAN networks, application licenses, NetScaler [Gateway] licenses, as well as the organization’s network infrastructure in general.

The Bottom Line

Citrix HDX Insight is a powerful yet cost-effective solution for obtaining unparalleled application and network-level visibility into XenDesktop and XenApp traffic on your network. With HDX Insight, IT teams can triage performance issues in real time, proactively monitor and manage developing situations before they balloon into bigger problems, and accurately plan for future server and network infrastructure needs.