Telegraf Integration Data Edit on GitHub Request doc changes

Contributors netapp-alavoie dgracenetapp

The Cloud Insights Telegraf data collectors are powerful tools for gathering source-reported metrics from your data sources.

Categorizing Telegraf Data

Because of the varied nature of the ingested data, Cloud Insights makes its best attempt to categorize the data in meaningful ways. To do this, it examines the meta-data (data about the data) that is gathers.

  • For meta-data that it knows, Cloud Insights attaches a meaningful name and categorizes the metric or counter as "Stable". Stable metrics will have names that use "Title Case", with individual words separated by space ( ), such as IOs In Progress.

  • For data that it might know but isn’t 100% certain, it keeps the source-determined name and assigns it to the "Guess" category. Guess metrics will appear with lower-case names, with individual words separated by an underscore (_), such as virtual_machine_io. "Guess" data cannot be used in filters or grouping.

All metrics gathered by Telegraf data collectors can be used in dashboard widgets. In the case of "Guess" metrics, you are given a choice of how to display the values for each metric you include (As-Is, Sum, Delta, etc.). These display options are explained below.

Widget example with Stable metrics:
Stable Metrics Example

Widget example with Guess metrics:
Guess Metrics Example

Telegraf Data Display Options

Cloud Insights provides different display options for Guess metrics collected using Telegraf. When adding these metrics to a widget, you are presented with a drop-down giving the data display choices:

As Is

Data is displayed as is, with no manipulation.

Rate

Current value divided by the time range since the previous observation.

Cumulative

The accumulation of the sum of previous values and the current value.

Delta

The difference between the previous observation value and the current value.

Delta rate

Delta value divided by the time range since the previous observation.

Cumulative Rate

Cumulative value divided by the time range since the previous observation.