Realtime performance alerts for cloud desktops

The thermostat for the first floor of my house is located right behind me in my home office. I probably walk by it a dozen times a day, but I only look at it if I am feeling cold or warm to check to see if the temperature is set to the correct value. In Massachusetts, hot and cold are seasonal but can vary a lot day to day so checking occurs more often than you would think. 

The modern thermostat has been around since Andrew Ure developed them in the 1830s. And in-person observation and control has been the primary use for this feedback mechanism for just as long. But what happens if we aren’t there to observe changes and take action?  Vacation homes and even primary residences where the owner is away for extended periods are at risk if heating components fail.  

In recent times, innovations such the Nest thermostat have combined the basic thermostat feedback loop with remote observation, using the Internet as the conduit to transport monitoring data to apps regardless of distance. These devices also continuously monitor the environment and can send text or email alerts if an abnormal condition such as room temperature falls below a threshold. 

Observing the behavior of cloud desktops and initiating appropriate troubleshooting and resolution steps require similar feedback loops.  Since Spot PC is based on Azure Virtual Desktop and Windows 365, the underlying Azure services and components typically experience high uptimes and consistent performance. However, no cloud solution is perfect and cloud desktops are Windows OS environments, which means they are susceptible to various combinations of application and resource failures. 

The Spot PC web console dashboard provides real time performance data that admins use to observe, diagnose, and resolve performance problems with cloud desktop session and virtual session machines. But what happens when admins are away from the console – how do they see that a session, virtual machine, or network may be in trouble?   

Starting this month, Spot PC is adding custom Monitors and Alerts to solve this problem. Monitors continuously observe one performance parameter – such as CPU utilization or network round trip time (RTT) – and creates an alert if the value matches a threshold value. For example, a monitor can be created to generate a Warning Alert if RTT exceeds 100 ms for more than 5 minutes or Critical Alert if RTT exceeds 200 ms for more than five minutes. Monitor rules can also generate an alert if a condition is met just once, such as a Windows service entering a Stopped state.

Monitors and alerts extend the observability features of Spot PC by continuously monitoring for abnormal conditions, even when admins are not logged into the console. Alerts can also be combined with logging information from the same time period to get more information on potential causes and failure sequences for specific problems. 

As the next step, alerts can also be sent to admin email addresses or applications that aggregate issues and alerts for administrators, such as PSA platforms like ConnectWise or Autotask or case systems like Zendesk or ServiceNow. In this way the alerts will be visible in centralized observability and admin task platforms, which in turn typically tie into active support workflows and can trigger appropriate responses. 

Spot PC is available today through Microsoft Azure Marketplace (for MSPs) and also by contacting the Spot PC team. To learn more, please visit the Spot PC website.