View a markdown version of this page

Time-series functions - Amazon CloudWatch Logs
offset
Services or capabilities described in Amazon Web Services documentation might vary by Region. To see the differences applicable to the China Regions, see Getting Started with Amazon Web Services in China (PDF).

Time-series functions

Use time-series functions with the stats command to analyze metrics over time windows and compute rates of change.

Function

Result type

Description

rate(fieldName: NumericLogField, interval: Period)

number

Computes the per-interval rate of change for a numeric field.

count_over_time(fieldName: LogField)

number

Counts log events per time bin. Use with by bin(interval) to set the window.

sum_over_time(fieldName: NumericLogField)

number

Sums field values per time bin. Use with by bin(interval) to set the window.

histogram(fieldName: NumericLogField, buckets: number)

map

Bucketizes numeric field values into the specified number of equal-width ranges and returns the distribution.

offset

Use offset at the end of a stats ... by bin() clause to shift time-series bins by a specified duration. This enables time-shifted comparisons, such as comparing current metrics against the same period in the previous hour or day.

Syntax

stats <aggregation> by bin(<period>) offset <duration>

Examples

stats count(*) by bin(5m) offset 1h
stats avg(latency) by bin(1m) offset 1d