Units of measurement

Observe works with many types of machine data which is presented in many different measurement units. This page discusses the units of measurement that can be used in Observe’s OPAL Console (via the set_metric verb) when using the Metrics Explorer or making Monitors, Dashboards, and Worksheets using metrics. Using set_metric or a visualization does not alter stored data, it only changes presentation.

When working with high precision evaluation of large scale metrics, it may be important to verify you are using the correct unit terms, such as Gibibytes instead of Gigabytes. Observe charts in Explorers, Dashboards, and Worksheets will scale charts by the unit size selected in Settings, Axes, Unit. For more information about charts, see Visualization types.

Use full unit names instead of abbreviations for better clarity.

Note that both names and abbreviations are case-sensitive.

📘

Note

Observe uses standard International System of Units (SI) names from the Math.js library, with the following exceptions:

  • The m abbreviation denotes minutes. Use the full name meter for length.
  • The C abbreviation denotes degrees celsius. Use the full name coulomb for electric charge.
  • The F abbreviation denotes degrees Fahrenheit. Use the full name farad for capacitance.\

Data units

Examples of data units:

Data unit nameAbbreviation
bitsb
bytesB
kilobytesKB
megabytesMB
gigabytesGB
terabytesTB
bytes/secondB/s
megabits/secondMb/s

Contact your Observe data engineer if you require help with an unusual or custom unit.

Percentages

This table shows how we handle abbreviations for percentages.

Data unit nameAbbreviations
Percent (0-100 range)
  • %
  • percent
  • % (0-100)
  • percent (0-100)
  • percentage
Percent (0.0-1.0 range)
  • % (0.0-1.0)
  • percent (0.0-1.0)
  • percentZeroToOne
  • ratio

Data ingestion terms

This table defines how data ingestion is measured for contracts and billing.

Data ingestDescription
GiBThe average number of gigabytes ingested per month.
Data Points per Minute (DPM)The average number of metric data points ingested per minute, typically measured on a rolling 30-day average.

A DPM is approximately the number of unique metric time series x reports per minute per series. A time series is the metric name + the full tag set.

Adding tags increases cardinality and DPM; adding more tag keys on the same series does not multiply DPM beyond defining that series.

For example:

  • A pod emits 1 cpu used percentage metric per minute = 1 DPM
  • The same metric emitted every 10s = 6 DPM
  • If the pod emits another metric with aws.region as a tag = 12 DPM
  • Every combination of tags emitted as metric on 10 second intervals would be additional 6 DPM

Did this page help you?