1aced0a130
as we will use distinct Influx and Grafana instances for the different stages/environments.
22 lines
948 B
Plaintext
22 lines
948 B
Plaintext
import "strings"
|
|
|
|
data = from(bucket: "poseidon")
|
|
|> range(start: v.timeRangeStart, stop: v.timeRangeStop)
|
|
|
|
runner_deletions = data
|
|
|> filter(fn: (r) => r["_measurement"] == "poseidon_used_runners")
|
|
|> filter(fn: (r) => r["event_type"] == "deletion")
|
|
|> keep(columns: ["_time", "id", "stage"])
|
|
|> rename(columns: {id: "runner_id"})
|
|
|
|
executions = data
|
|
|> filter(fn: (r) => r["_measurement"] == "poseidon_nomad_executions" or r["_measurement"] == "poseidon_aws_executions")
|
|
|> filter(fn: (r) => r["event_type"] == "creation")
|
|
|> filter(fn: (r) => contains(value: r["environment_id"], set: ${environment_ids:json}))
|
|
|> keep(columns: ["_value", "environment_id", "runner_id"])
|
|
|> count()
|
|
|
|
result = join(tables: {key1: executions, key2: runner_deletions}, on: ["runner_id"], method: "inner")
|
|
|> keep(columns: ["_value", "_time", "environment_id", "stage"])
|
|
|> aggregateWindow(every: v.windowPeriod, fn: mean, createEmpty: false)
|