I am currently querying data from a Parquet data lake stored in Azure Data Lake Gen2 via Grafana using Microsoft SQL Server as my data source. My query looks as below:

SELECT
    DATEADD(SECOND, DATEDIFF(SECOND, '2020', t) / ($__interval_ms/1000) * ($__interval_ms/1000), '2020') AS time,
    AVG(accelerationx) as AVG_accelerationx, AVG(accelerationy) as AVG_accelerationy, AVG(accelerationz) as AVG_accelerationz
FROM
    OPENROWSET(
         BULK 'https://cssdatalakestoragegen2.dfs.core.windows.net/cssdatalakestoragegen2filesystem/3BA199E2/CAN2_gnssimu/*/*/*/*',
         FORMAT = 'PARQUET'
   ) AS r
WHERE
    CONCAT(r.filepath(1), '-', r.filepath(2), '-', r.filepath(3)) BETWEEN '2020-10-28' AND '2020-10-28'
    AND t BETWEEN '2020-10-28 14:44:00' AND '2020-10-28 14:48:00'
GROUP BY
    DATEDIFF(SECOND, '2020', t) /  ($__interval_ms/1000)
ORDER BY time
OFFSET 0 ROWS;

This works when the value of $__interval_ms is 1000 or larger, but fails for smaller values. This is a problem as I need to be able to zoom in/out of my plot dynamically – switching between grouping data points e.g. at 5 min intervals to 1 ms intervals.

Is there a way to tweak the query so that I achieve the above? I seem to encounter errors related to overflow and other issues when doing modifications to this.