hardware-metrics

Uptime %

  • Used to measure uptime / availability.
  • Issues with this metric equate to reduced revenue generation.

Response Time ms

  • Measurement of how long it takes the application or infrastructure to respond.
  • Poor response time indicates communication or processing issues along the application path.

Error Rate %

  • Percentage of errors generated per minute.
  • Higher error rates indicate problems with the application or the hosted infrastructure.

Requests per Second

  • Indication of incoming load to the environment; precursor to other load indicators.
  • Higher numbers mean that the environment will start to experience heavy load.

Fast, Slow, Very Slow, Stalled, Failed Requests

  • Indicates how Application Transactions are performing.
  • Poor numbers indicate an issue with a transaction or the hardware supporting the application.

Load

  • How efficiently the node is using CPU, Memory, and Disk.
  • Load ≥ 8 indicates the node is having difficulty keeping up with work.

Open Files Descriptors (OFD)

  • Measurement of the current/total available OFD.
  • High percentage of OFD indicates an unhealthy application.

Read more: