Zero-Filling not working for non available buckets.

Affected Version : 24.0

### Description

While trying to query in a interval (example: "intervals": [ "2021-01-01T00:00:00.000/2022-01-04T00:00:00.000" ])

The buckets which are empty (which do not have data matching intervals with the data of timestamp in the database) does not populate zero filling even though keeping skipEmptyBuckets = false.

Query using:

{
"queryType": "timeseries",
"dataSource": "sample2",
"granularity": "day",
"aggregations": [
{ "type": "longSum", "name": "sample_name1", "fieldName": "content.VOLUME" }
],
"intervals": [ "2021-01-01T00:00:00.000/2022-01-04T00:00:00.000" ],
"context" : {
"skipEmptyBuckets": "false"
}
}

Not sure if this is not possible or if its an error?

Start with Druid's timeseries query handling and the skipEmptyBuckets context option, using the supplied interval and query as the reproduction. Confirm whether empty day buckets are omitted despite skipEmptyBuckets=false, then identify the relevant behavior or test and define done as returning zero-filled buckets for the requested interval, or documenting that this is unsupported.

Written by the indexing model from the issue text.