Breakdown: Troubleshooting Issues with VMware vCenter Operations Manager for Horizon View
By Cindy Heyer, Contract Technical Writer for Technical Marketing, End-User Computing at VMware; with significant contributions from David Wooten, Product Manager for VMware vCenter Operations Manager for Horizon View; and David Homoki, Senior Member of the Technical Staff at VMware
Breakdowns happen. And the Deployment Guide for VMware vCenter Operations Manager for Horizon View provides solutions to breakdowns that happen in communication between the Adapter, the Broker Agent, and the Desktop Agents–the three main components of the VMware vCenter Operations Manager for Horizon View (vC Ops for View). VMware vC Ops for View is an extension of Horizon vCenter Operations Manager. If issues occur during installation or configuration, the result can show up as missing metric values, blue Health Badges in the object tree, etc. The vC Ops for View Deployment Guide describes how to resolve issues like these.
The vC Ops for View Deployment Guide is the first such guide for this product. This guide provides IT administrators with helpful tips and typical scenarios for troubleshooting your vC Ops for View instance as you use it to monitor and troubleshoot your network infrastructure. If you want to add your input to our list of use case scenarios, send your suggestions to twitter.com/vmwarehorizon.
Is the Adapter Receiving Data from the Broker Agent?
An example of a typical question that the Deployment Guide covers is the communication between the Adapter and the Broker Agent. The Broker Agent Status Widget should be populated with valid collection times and counts, and the health score of the Broker Agent should not be blue.
By default, the graphs show a 6-hour time period, and should extend all the way to the right side of each box. If this is not the case, it could be due to a breakdown of communication between the Adapter and Broker Agent.
The vC Ops for View Deployment Guide provides a checklist to help you systematically verify that the vC Ops for View Adapter is receiving data from your Broker Agent properly:
1. Check the health score. Look at the health score of the Broker Agent in the View Adapter Status Dashboard, and verify that the Broker Agent is configured properly to send data to the Adapter:
a. Using the configuration utility, verify that pairing is successful.
b. Use the Test button to verify connectivity with the View Adapter.
2. Check the Broker Agent log. Locate the Broker Agent log (by default, at ProgramData\VMware\vCenter Operations for View\logs), and search in GetTopology task results for the following line, with appropriate values for your environment:
GetTopology task results: Pods 1; VCs 1; Brokers 1; Pools 6; Desktop VMs 9; Sessions 1
If the session count is 0, but the View Administrator shows sessions, restart the Broker Agent.
A problem can occur when the View Connection Server where the Broker Agent is running gets rebooted. When the View Connection Server restarts, the Broker Agent starts before all Horizon View services begin running. This causes the Broker Agent to be in a bad state. Restarting the Broker Agent clears this state.
3. Check for a timeout issue. In larger environments, the PowerShell call to collect the session data can take a long time. The call has a timeout of five minutes. If it takes longer than that, the Broker Agent will not report session data to the Adapter. You can check from either the Adapter or the Broker Agent:
a. From the Adapter – Select the Broker Agent in the Broker Agent Status widget, expand the Broker Agent Topology Collection Statistics widget, and check the User Session Collection Time. If the collection time is greater than 300 seconds and the User Session Count is 0, the call may be timing out.
b. From the Broker Agent – You may see something like the following:
Timed out waiting for task to complete: TaskGetViewRemoteSession.
Alternatively, you can search for the most recent line containing
GetTopology task timing
This line shows the time taken to collect each part of the topology, including the sessions. If the time is greater than 300 seconds, this may indicate that a timeout is occurring. As a confirmation, the next line in the log should be
GetTopology task results
This line indicates the number of sessions collected. If the time is greater than 300 seconds, and the number of sessions in this line is 0, the call is timing out.
This is just one example of the tips and troubleshooting scenarios included in the vC Ops for View Deployment Guide. The Deployment Guide covers other issues that can occur during installation and configuration, such as
- Login or reconnect times fail to display
- PCoIP metrics do not display properly
- PowerShell errors appear in the Broker Agent log
See the vC Ops for View Deployment Guide for additional use case scenarios.
You can also