New Grafana user here, so go easy on me!
Desired behaviour: When I select a host and cycleNumber (both variables), panels populate with data collected between a given window of time during that cycle. This window of time is defined by “Step”, which increments during each cycle. I’m interested in data between steps 7 and 44.
To do this, I think I need to:
- Get the time when cycleNumber changed to the desired value
- Get the first time AFTER cycleNumber time when step == 7
- Get the first time AFTER cycleNumber time when step == 44
- Use the times of step == 7 and 44 to define a range to query other desired data (temperatures, valve states etc)
Step 1 is completed successfully using:
from(bucket: "my_bucket")
|> range(start: 0) // Query from the first available data
|> filter(fn: (r) => r["_measurement"] == "data")
|> filter(fn: (r) => r["host"] =~ /^$host$/) // Taken from a variable
|> filter(fn: (r) => r["_field"] == "CycleNumber" and r["_value"] == ${CycleNumber_variable}) // Also taken from a variable
|> keep(columns: ["_time"])
Now, I need to either store this time in a variable or pipe it into queries for steps 2,3 and 4.
Failed attempts at using in a variable:
- Added to above query
|> findRecord(fn: (key) => true, idx: 0) // Extract the time as a record |> findColumn(fn: (key) => true, column: "_time")-
|> range(start: 0) // Adjust as needed |> filter(fn: (r) => r._measurement == "data") |> filter(fn: (r) => r.host == "${host}") |> filter(fn: (r) => r._field == "CycleNumber") |> filter(fn: (r) => r._value == ${CycleNumber_variable}) |> sort(columns: ["_time"], desc: false) |> limit(n: 1) |> map(fn: (r) => ({_value: r._time}))```
The above and various permutations don’t populate a variable with any value. Is there sometime obvious I’m missing?
Other approach is to try to use blocks in the same query, the first defining a namespace to be used in the second. For example:
time_of_switch = from(bucket: "my_bucket")
|> range(start: 0)
|> filter(fn: (r) => r._measurement == "data")
|> filter(fn: (r) => r.host == "${host}")
|> filter(fn: (r) => r._field == "cycle_number")
|> filter(fn: (r) => r._value == int(v: "${cycle_number_variable}"))
|> sort(columns: ["_time"], desc: false)
|> limit(n: 1)
|> keep(columns: ["_time"])
|> findColumn(fn: (key) => true, column: "_time")
// Step 2: Find the first instance where Step equals 7 after the time_of_switch
from(bucket: "my_bucket")
|> range(start: time(v: time_of_switch))
|> filter(fn: (r) => r._measurement == "data")
|> filter(fn: (r) => r._field == "Step")
|> filter(fn: (r) => r.host == "${host}")
|> filter(fn: (r) => r._value == 7)
|> first()
I’ve also tried various permutations of this, but this only displays the first timestamp from the first block.
Am I barking up the wrong tree or am I close?
