Windows Insider
How To Troubleshoot with the RDS Connection Quality Indicator
Here's how to take advantage of Microsoft Remote Desktop Services (RDS) 8.0's new Connection Info notification in the Remote Desktop Client.
- By Greg Shields
- 04/28/2014
Sometimes I feel bad for technologies such as Microsoft Remote Desktop Services (RDS). When users have problems with remote applications, all too often they blame RDS.
You've surely experienced this misrepresentation yourself. How many times have your users called in with "a Remote Desktop problem" that was really just a bad network connection?
Until recently, the options for proving RDS wasn't at fault have been relatively limited. You could ask the user to ping a server and then read off whatever latency values he found. Doing so isn't a great test, but it's better than nothing. More often than not, your troubleshooting process probably required asking a few subjective questions about the quality of network the user was connecting from.
"A hotel today? An airport, perhaps? Stuck with just 3G connectivity? Maybe that's your problem."
Smarter Client
RDS 8.0 adds a new Connection Info notification to the Remote Desktop Client (RDC) that should help with many of these finger-pointing phone calls. Much like the signal bars on your cell phone, the new connection quality indicator in RDS 8.0 is designed to let users know when the network might be the problem.
You'll find the RDC indicator up in the connection bar at the top of any full-screen remote desktop connection. The indicator will display between one and four "bars" that summarize the connection's roundtrip time (RTT) latency and current bandwidth into an easy-to-read icon.
A full set of four bars, for example, indicates the latency is less than 50 ms and the bandwidth is greater than 10 Mbps. When there are fewer bars the user is on a higher latency or lower bandwidth connection. Microsoft outlines the relationship between the number of bars shown and the associated measurements in the knowledgebase article, "Introduction to the Connection Info feature in the RDC client in Windows 8 and Windows Server 2012".
A user can also click the indicator's icon to see a written description of his connection's quality and to learn if he's enjoying the added benefits of a UDP-based connection.
Unfortunately, this version's RDC only displays that connection quality indicator when the remote connection is made in full-screen mode. Windowed connections or those to RemoteApps get no such visual notification for end users. The Connection Info feature is also available when connections are made via the Remote Desktop app from the Windows Store.
While you might hope Microsoft adds this visual notification to all connection types in future RDC versions, there is a server-side workaround that'll deliver you the administrator access to the information.
Take a look at Performance Monitor (PerfMon) on a server running RDP 8.0. The four counters you're looking for can be found under the RemoteFX Network category: Current TCP RTT, Current UDP RTT, Current TCP Bandwidth and Current UDP Bandwidth. As is obvious, these four identify the current bandwidth and roundtrip time latency per RDP session.
The graph view is less valuable here than the report view, which offers an instantaneous value for bandwidth and RTT per session. The extra step here is in translating the RDP-TCP session number to an actual human being, a process that as of this writing still requires the ancient Terminal Services query user command. You'll need to run query user on the server in question to get the mapping between a logged-on username and the RDP-TCP session number.
Bad Network!
The fact that these values are stored in PerfMon makes their use in a scripted solution entirely possible with a little work. The Windows PowerShell Get-Counter command against those four metrics in the RemoteFX Network(*) category can quickly spit out a report that can be massaged into a kind of RDS connection-quality dashboard.
Ultimately, these and the other metrics in the RemoteFX Network category are designed to help you -- the troubleshooting administrator -- track down the real cause for remote application issues. With these metrics in hand, a bad network connection need no longer be an invisible enemy.
About the Author
Greg Shields is Author Evangelist with PluralSight, and is a globally-recognized expert on systems management, virtualization, and cloud technologies. A multiple-year recipient of the Microsoft MVP, VMware vExpert, and Citrix CTP awards, Greg is a contributing editor for Redmond Magazine and Virtualization Review Magazine, and is a frequent speaker at IT conferences worldwide. Reach him on Twitter at @concentratedgreg.