Long Running Query after 1.8 -> 2

InfluxDB 2
1

Trying to migrate all my metrics from influx 1.8 to 2.x and have one query that just seems to take forever.

I’m trying to track the % used on 16 HDDs for my NAS over a 30 day range. In 1.8, it ran just fine and never timed out. Now, it seems to take almost 30s just to run.

Here’s the actual flux query:

from(bucket: "armin")
  |> range(start: v.timeRangeStart, stop: v.timeRangeStop)
  |> filter(fn: (r) => r["_measurement"] == "disk")
  |> filter(fn: (r) => r["_field"] == "used_percent")
  |> filter(fn: (r) => r["device"] == "mergerfs" or r["device"] == "sda" or r["device"] == "sdb" or r["device"] == "sdc" or r["device"] == "sdd" or r["device"] == "sde" or r["device"] == "sdf" or r["device"] == "sdg" or r["device"] == "sdi" or r["device"] == "sdh" or r["device"] == "sdj" or r["device"] == "sdk" or r["device"] == "sdl" or r["device"] == "sdm" or r["device"] == "sdn" or r["device"] == "sdo")
  |> group(columns: ["device"])
  |> aggregateWindow(every: v.windowPeriod, fn: last, createEmpty: false)
  |> yield(name: "last")

It works and returns the data that I’m looking for, but it takes almost 25 seconds to run in the Influx Data Explorer. Takes even longer when running via Grafana.

I’m sure there’s a simple tweak I can do to make it a lot more efficient, just not sure as I’m learning flux as I go.

2

Hello @kyleinprogress,
I actually recommend waiting to upgrade until 3.x OSS is available later this year.
That’s because it supports InfluxQL so itll be easier to migrate.
Additionally you get these performance benefits:

Its not surprising to me, Flux has some issues. :frowning: I’m sorry I know this isn’t the response you were hoping for.

3

@kyleinprogress your only other option is to include this type of query in a task that continuously writes the output to a new bucket so you can query the transformed data directly and perform it with smaller chunks of data.