Simple Log Service:PromQL function usage examples

abs()

Returns the absolute value of each data point in the input time series.

abs(container_network_receive_bytes_total{})

changes()

Returns the number of times each time series value changes within the specified time range.

changes(demo_api_request_duration_seconds_count[1m])

absent()

Returns an empty vector if the input vector has any elements. Returns a 1-element vector with the value 1 if the input vector has no elements. Use this function to alert when a metric has no data.

absent(test_metrics)

absent_over_time()

Returns 1 if there are no data points within the specified time range. Returns an empty vector otherwise. Use this function to detect gaps in metric data over a range.

absent_over_time(test_metrics[1m])

ceil()

Rounds each data point up to the nearest integer.

ceil(process_start_time_seconds)

clamp()

Clamps each data point to a specified range. Values below min are set to min; values above max are set to max.

Syntax: clamp(v instant-vector, min scalar, max scalar)

clamp(process_start_time_seconds,10,20)

clamp_max()

Clamps each data point to a maximum value. Values above max are set to max. Similar to clamp(), but applies an upper bound only.

Syntax: clamp_max(v instant-vector, max scalar)

clamp_max(process_start_time_seconds,10)

clamp_min()

Clamps each data point to a minimum value. Values below min are set to min. Similar to clamp(), but applies a lower bound only.

Syntax: clamp_min(v instant-vector, min scalar)

clamp_min(process_start_time_seconds,10)

day_of_month()

Returns the day of the month (1–31) for each timestamp in the input time series.

day_of_month(vector(1805308033))

day_of_week()

Returns the day of the week (0–6, where 0 is Sunday) for each timestamp in the input time series.

day_of_week(vector(1805308033))

day_of_year()

Returns the day of the year (1–365) for each timestamp in the input time series.

day_of_year(vector(1805308033))

delta()

Returns the difference between the first and last data points within the specified time range. Use with gauges.

delta(go_memstats_heap_released_bytes[1m])

deriv()

Returns the per-second derivative of each time series in the range vector, calculated using simple linear regression. Use with gauges.

deriv(go_memstats_heap_released_bytes[1m])

exp()

Returns the exponential function e^x for each data point in the input time series.

exp(go_memstats_lookups_total)

floor()

Rounds each data point down to the nearest integer.

floor(go_memstats_heap_released_bytes)

holt_winters()

Produces a smoothed value for each time series using double exponential smoothing (Holt-Winters algorithm). The smoothing factor sf controls how much weight is given to older data (lower = more weight on old data); the trend factor tf controls how strongly trends are followed. Both sf and tf must be between 0 and 1. Use with gauges.

Syntax: holt_winters(v range-vector, sf scalar, tf scalar)

holt_winters(go_memstats_heap_released_bytes[3m],0.2,0.5)

rate()

Returns the per-second average rate of increase of the counter over the specified time range. Accounts for counter resets. Use with counters.

rate(go_memstats_mallocs_total[3m])

idelta()

Returns the difference between the last two data points within the specified time range.

idelta(go_memstats_mallocs_total[3m])

increase()

Returns the total increase of a counter over the specified time range. Counter resets (where the value drops to zero and resumes counting) are automatically accounted for. Use with counters only.

increase(go_memstats_mallocs_total[3m])

irate()

Returns the per-second instantaneous rate of increase, calculated from the last two data points in the range. Use with fast-moving counters and for graphing. For alerting or slow-moving counters, use rate() instead.

irate(go_memstats_mallocs_total[3m])

hour()

Returns the hour of the day (0–23 UTC) for each timestamp in the input time series.

hour(vector[1705308033])

minute()

Returns the minute of the hour (0–59 UTC) for each timestamp in the input time series.

minute(vector[1705308033])

month()

Returns the month of the year (1–12 UTC) for each timestamp in the input time series.

month(vector[1705308033])

resets()

Returns the number of counter resets within the specified time range. A reset occurs when the counter value drops between two consecutive data points. Use with counters.

resets(go_memstats_heap_sys_bytes[5m])

round()

Rounds each data point to the nearest multiple of to_nearest. If to_nearest is omitted, rounds to the nearest integer.

Syntax: round(v instant-vector, to_nearest=1 scalar)

round(go_memstats_heap_sys_bytes)