InfluxDB: Flux - Aggregation Function reduce(): Unterschied zwischen den Versionen

Aktuelle Version vom 19. Mai 2021, 13:08 Uhr

When should I use reduce()

The aggregation function "reduce()" is useful if you want to aggregate (sum, count,..) foreach series.
For example you have multiple Clusters with multiple Nodes in it and you want to calculate how many Nodes are in each Cluster and also calculate the sum of a "_value" field, then you can use reduce().
I think it is more clearly when you have read the example below.

Important

By default, reduce() drops any columns that:

• Are not part of the input table’s group key.
• Are not explicitly mapped in the reduce() function.

Example

Because the raw data in my example measurement "asset_hyperv_local" has a lot of tags I use "keep()" to decrease the number of columns.

Flux

bucket = "<YOUR_BUCKET>"

from(bucket: bucket)
  |> range(start: v.timeRangeStart, stop: v.timeRangeStop)
  |> filter(fn: (r) => 
    r._measurement == "asset_hyperv_local" and
    r._field == "physicalmemory"
  )
  |> last()
  |> group(columns: ["cluster", "manufacturer"])
  |> keep(columns: ["_time", "_value", "cluster", "manufacturer", "_field"])

Same Flux Query but with "reduce()" in action

bucket = "<YOUR_BUCKET>"

from(bucket: bucket)
  |> range(start: v.timeRangeStart, stop: v.timeRangeStop)
  |> filter(fn: (r) => 
    r._measurement == "asset_hyperv_local" and
    r._field == "physicalmemory"
  )
  |> last()
  |> group(columns: ["cluster", "manufacturer"])
  |> keep(columns: ["_time", "_value", "cluster", "manufacturer", "_field"])
  |> reduce(
    fn: (r, accumulator) => ({ 
      NumberOfNodes: accumulator.NumberOfNodes + 1.0,
      SumOfPhysicalMemory: r._value + accumulator.SumOfPhysicalMemory,
    }), 
    identity: {NumberOfNodes: 0.0, SumOfPhysicalMemory: 0.0}
  )

Influxdata documentation: https://docs.influxdata.com/influxdb/v2.0/reference/flux/stdlib/built-in/transformations/aggregates/reduce/