Service-level monitoring is possible for Connection Servers and Unified Access Gateways.Configuring telegraf enables you to monitor Tomcat services as well as perform HTTP Health Checks against the UAG and CS web services. 20-sec monitoring metrics can be enabled in vCenter Adapter for Horizon use cases.

Following three configurations can be used:

  1. Configuring Telegraf :

    Install the telegraf agent on the CS Server by navigating through Inventory → Manage Telegraf Agents. Telegraf Agent Installation

    • Navigate to Environment -> Applications -> Manage Telegraf Agents page

    • Select the UAG / CS Server click the horizontal ellipsis, and then click Install

    • Provide the server administrator user credentials and install the telegraf.

    Note:

    Telegraf installation for UAG servers with version 21 and above fails and the client receives SOAP fault from server. As a pre-requisite, add a "tmp" folder with "exec" permission. Use the following command and retry installation.<mount -o remount,exec /tmp>

    Sample Configurations for UAG Server:

    1. Tomcat

      Base URL : http://localhost:8877/

      Installed Path : /opt/vmware/gateway/data/authbroker/.extractOR /opt/vmware/gateway/.extract

      Note:

      Install Path is where the horizon client is deployed over UAG server. Precisely where hc.war is extracted

    2. HttpCheck :

      URL : https://localhost:443/

      Method : GET

      Skip host and chain verification :true

    3. PingCheck

      DisplayName :

      FQDN / IP : Provide UAG Server IP

    Sample Configurations for Connection Server:

    1. Tomcat

      Base URL : http://localhost:48080

      Installed Path : C:\Program Files\VMware\VMware View\Server\broker\

    2. HttpCheck :

      URL : http://localhost:48080

      Method : GET

    3. PingCheck

      DisplayName :

      FQDN / IP : Provide CS Server IP

  2. Enabling 20-sec Monitoring :

    Navigate to Configure > Policies, and then click vSphere Solutions Default Policy.

    1. Click Edit Policy to enable peak metrics for the Virtual Machine.

    2. Click Metrics & Properties, and then select the object type as Virtual Machine.

    3. Expand Metrics section and edit the state to enable below metrics:

      1. Under CPU

      2. Peak vCPU Co-Stop within collection cycle

      3. Peak vCPU IO wait

      4. Peak vCPU overlap

      5. Peak vCPU Ready

      6. Peak vCPU Swap Wait

      7. Peak vCPU System

      8. Under Memory->Peak Contention

      9. Under Guest

      10. Peak CPU Context Switch

      11. Peak vCPU Queue

      12. Peak vCPU Disk Queue

      13. Under Network

      14. Peak Usage Rate

      15. Virtual Disk

      16. Peak Latency

  3. Configuring Unified Access Gateways :

    Host name of UAG VM, DNS name, and Unified Access Gateway name provided while deploying the UAG must be in the same order to monitor the metrics & sessions.