Grafana v10.2.6 (NA)
I have a influxdb instance with a table where one of the tags are called “bucket”. I can use it in other fluxql functions like sort
, map
etc. But when trying to use it in filter
it is replace by ""
Entire query
from(bucket: "fixed_telemetry_historic")
|> range(start: v.timeRangeStart, stop: v.timeRangeStop)
|> filter(fn: (r) => r._measurement == "nuttcp_distrubution"
and r._field == "bucket_mbps"
and r.direction == "down"
and r.bucket == "1" // r.bucket in use
)
|> sort(columns:["bucket"])
From the querry inspector
from(bucket: "fixed_telemetry_historic")
|> range(start: 2025-05-20T09:44:22.878Z, stop: 2025-05-21T09:44:22.878Z)
|> filter(fn: (r) => r._measurement == "nuttcp_distrubution"
and r._field == "bucket_mbps"
and r.direction == "down"
and "" == "1" // wait? Where did my r.bucket go?
)
|> sort(columns:["bucket"])
When I run the same querry in influxdb’s web interface it returns the expected result from the table.
Is r.bucket
reserved in grafana? If so it should be (better) documented. And it should produce an error in the UI so it is clear to the user. If it is not reserved, it seems I have stumbled across a bug which should be fixed.