Hey there! Let's get that count-by-value query working for you. Your initial approach of using COUNT(*) with GROUP BY is on the right track—you just need to tweak the syntax to match InfluxQL's requirements, and add a few touches to get the exact output format you want.

The Correct InfluxQL Query #

Here's the query that will give you the count of results per foo value:

SELECT COUNT(*) AS numResults FROM myMeasurement GROUP BY "foo"

Breaking Down Why This Works #

• COUNT(*): This counts every data point in each group, which is exactly what you need to tally how many times each foo value appears.
• AS numResults: Renames the count column to match the numResults name you want in your output.
• GROUP BY "foo": Groups your data by the distinct values in the foo field. Wrapping foo in double quotes is a safe practice in InfluxQL—it avoids conflicts with keywords and ensures the field name is interpreted correctly.

What the Output Will Look Like #

When you run this query, InfluxDB will return results grouped by each foo value (formatted as series with tags):

name: myMeasurement
tags: foo=valOne
time                 numResults
----                 ----------
1970-01-01T00:00:00Z 3

name: myMeasurement
tags: foo=valTwo
time                 numResults
----                 ----------
1970-01-01T00:00:00Z 2

name: myMeasurement
tags: foo=valThree
time                 numResults
----                 ----------
1970-01-01T00:00:00Z 4

If You Want a Compact Table Format (InfluxDB 2.x+) #

If you're using InfluxDB 2.x, Flux (the newer query language) can return results in the exact tabular format you showed. Here's the Flux query:

from(bucket: "your-bucket-name")
  |> range(start: 0) // Adjust start time if you don't want all historical data
  |> filter(fn: (r) => r._measurement == "myMeasurement" and r._field == "foo")
  |> group(columns: ["_value"])
  |> count(column: "_value")
  |> rename(columns: {_value: "foo", count: "numResults"})
  |> keep(columns: ["foo", "numResults"])

This will output:

foo        numResults
---        ----------
valOne     3
valTwo     2
valThree   4

内容的提问来源于stack exchange，提问作者ekjcfn3902039