FAQs about telegraf agents

Products

• VCF Operations

Issue/Introduction

Question:

If Aria Operations Telegraf agent is installed on the VM and OS is detected and Aria Operations shows the Windows host or Linux host configured:

• What exactly is being monitored by this telegraf agent configuration out of the box?
• What are the alert definition for this kind of alerts?
• Where can what is monitored be observed?

Environment

Aria Operations 8.18.x

Resolution

Answer:

• When the Telegraf agent is installed and either the "Windows OS" or "Linux OS" object is detected, the agent defaults to monitoring OS-level performance metrics. Unlike the standard vSphere VM metrics (which are observed from the outside by the hypervisor), these metrics are observed from inside the Guest OS. The default metrics monitored are given below:
  • CPU: Usage breakdown (User, System, Idle, Wait), Context switches.
  • Memory: Active, Available, Buffered, Cached, Free, Used, Paging/Swap statistics.
  • Disk/Filesystem:
    • Usage: Used Space (GB), Free Space (GB), Usage Percentage (%) per mount point/drive (e.g., C:\, /, /boot).
    • I/O: Read/Write bytes, IOPS.
  • Network: Bytes Sent/Received, Packets Sent/Received, Errors, Drops per interface.
  • Agent Health: Status of the Telegraf service itself (Up/Down).
    Note: Specific applications (e.g., SQL, IIS, Apache) or individual processes/services are not monitored out-of-the-box unless you explicitly enable a specific "Service" configuration or "Process Monitor" within the agent settings. As referenced in the document: OS and Application Monitoring.
• To view the exact list of metrics currently being ingested for a specific VM:
  
  • Navigate to Inventory > vSphere Host and Cluster.
  • Expand the hierarchy or search for the VM name.
  • Look for the child object named "Windows OS" or "Linux OS" (this is distinct from the "Virtual Machine" object).
  • Select the OS object and click the Metrics tab.
    
    • This view lists every single data point the agent is successfully collecting.
      As referenced in the document: OS and Application Monitoring.
• Alert definitions for these agents are distinct from vSphere alerts because they target the TelegrafAdapter object types. To locate the definitions:
  
  • Navigate to Operations > Configure > Alerts > Alert Definitions.
  • Click the Filter icon.
  • Set Adapter Type to TelegrafAdapter (or Application Monitoring depending on the version).
  • (Optional) Set Object Type to Windows OS or Linux OS.
    As referenced in the document: Alert Definitions in VMware Aria Operations.

For some addition common Out-of-the-Box Alerts courtesy of document: Alert Definitions in VMware Aria Operations, please see the below:

• Telegraf Agent is Down: Triggered if the agent stops sending heartbeats.
• Filesystem Usage High: Triggered if a specific Guest OS partition exceeds usage thresholds (e.g., >90%).
  
  • Note: This is often more accurate than vSphere-level "Guest File System" alerts because it sees all mount points.
• High CPU Usage (OS Level): Triggered based on internal OS counters rather than hypervisor demand.

Additional Information

OS and Application Monitoring

 Alert Definitions in VMware Aria Operations.