Documentation

Query using conditional logic

This page documents an earlier version of InfluxDB. InfluxDB v1.8 is the latest stable version. View this page in the v1.8 documentation.

See the equivalent InfluxDB v2.0 documentation: Query using conditional logic.

Flux provides if, then, and else conditional expressions that allow for powerful and flexible Flux queries.

Conditional expression syntax
// Pattern
if <condition> then <action> else <alternative-action>

// Example
if color == "green" then "008000" else "ffffff"

Conditional expressions are most useful in the following contexts:

  • When defining variables.
  • When using functions that operate on a single row at a time ( filter(), map(), reduce() ).

Evaluating conditional expressions

Flux evaluates statements in order and stops evaluating once a condition matches.

For example, given the following statement:

if r._value > 95.0000001 and r._value <= 100.0 then "critical"
else if r._value > 85.0000001 and r._value <= 95.0 then "warning"
else if r._value > 70.0000001 and r._value <= 85.0 then "high"
else "normal"

When r._value is 96, the output is “critical” and the remaining conditions are not evaluated.

Examples

Conditionally set the value of a variable

The following example sets the overdue variable based on the dueDate variable’s relation to now().

dueDate = 2019-05-01
overdue = if dueDate < now() then true else false

Create conditional filters

The following example uses an example metric variable to change how the query filters data. metric has three possible values:

  • Memory
  • CPU
  • Disk

    metric = "Memory"
    
    from(bucket: "telegraf/autogen")
    |> range(start: -1h)
    |> filter(fn: (r) =>
      if v.metric == "Memory"
        then r._measurement == "mem" and r._field == "used_percent"
      else if v.metric == "CPU"
        then r._measurement == "cpu" and r._field == "usage_user"
      else if v.metric == "Disk"
        then r._measurement == "disk" and r._field == "used_percent"
      else r._measurement != ""
    )
    

Conditionally transform column values with map()

The following example uses the map() function to conditionally transform column values. It sets the level column to a specific string based on _value column.

from(bucket: "telegraf/autogen")
  |> range(start: -5m)
  |> filter(fn: (r) => r._measurement == "mem" and r._field == "used_percent" )
  |> map(fn: (r) => ({
    r with
    level:
      if r._value >= 95.0000001 and r._value <= 100.0 then "critical"
      else if r._value >= 85.0000001 and r._value <= 95.0 then "warning"
      else if r._value >= 70.0000001 and r._value <= 85.0 then "high"
      else "normal"
    })
  )
from(bucket: "telegraf/autogen")
  |> range(start: -5m)
  |> filter(fn: (r) => r._measurement == "mem" and r._field == "used_percent" )
  |> map(fn: (r) => ({
    // Retain all existing columns in the mapped row
    r with
    // Set the level column value based on the _value column
    level:
      if r._value >= 95.0000001 and r._value <= 100.0 then "critical"
      else if r._value >= 85.0000001 and r._value <= 95.0 then "warning"
      else if r._value >= 70.0000001 and r._value <= 85.0 then "high"
      else "normal"
    })
  )

Conditionally increment a count with reduce()

The following example uses the aggregateWindow() and reduce() functions to count the number of records in every five minute window that exceed a defined threshold.

threshold = 65.0

from(bucket: "telegraf/autogen")
  |> range(start: -1h)
  |> filter(fn: (r) => r._measurement == "mem" and r._field == "used_percent" )
  |> aggregateWindow(
      every: 5m,
      fn: (column, tables=<-) => tables |> reduce(
            identity: {above_threshold_count: 0.0},
            fn: (r, accumulator) => ({
              above_threshold_count:
                if r._value >= threshold then accumulator.above_threshold_count + 1.0
                else accumulator.above_threshold_count + 0.0
            })
        )
    )
threshold = 65.0

from(bucket: "telegraf/autogen")
  |> range(start: -1h)
  |> filter(fn: (r) => r._measurement == "mem" and r._field == "used_percent" )
  // Aggregate data into 5 minute windows using a custom reduce() function
  |> aggregateWindow(
      every: 5m,
      // Use a custom function in the fn parameter.
      // The aggregateWindow fn parameter requires 'column' and 'tables' parameters.
      fn: (column, tables=<-) => tables |> reduce(
            identity: {above_threshold_count: 0.0},
            fn: (r, accumulator) => ({
              // Conditionally increment above_threshold_count if
              // r.value exceeds the threshold
              above_threshold_count:
                if r._value >= threshold then accumulator.above_threshold_count + 1.0
                else accumulator.above_threshold_count + 0.0
            })
        )
    )

New! Cloud or OSS?