Apache Pinot

say if your query is `select sum(met0) from myTable group by dim0`

<@U012LRYSXNH> If you are looking for faster latency for these queries, try enabling star-tree index

<@UDRJ7G85T> if i enable it i should re-create the entire table right ?

Yes. There is a PR to create it on the fly without having to reload data

not a big problem because as you told me when i run again the command:

```docker run --rm -ti 
	--network=pinot-demo
	-v /tmp/pinot-quick-start:/tmp/pinot-quick-start
	--name pinot-data-ingestion-job
	apachepinot/pinot:latest LaunchDataIngestionJob
	-jobSpecFile /tmp/pinot-quick-start/docker-job-spec.yml```


the entire table will be overwritten again

<@UDRJ7G85T> i try to find on the doc how to activate that index type

```{
  "tableName": "test",
  "tableType": "OFFLINE",
  "segmentsConfig": {
    "segmentPushType": "REFRESH",
    "segmentAssignmentStrategy": "BalanceNumSegmentAssignmentStrategy",
    "replication": "1"
  },
  "tenants": {},
  "tableIndexConfig": {
    "loadMode": "MMAP"
  },
  "metadata": {
    "customConfigs": {}
  }
}```


so my query is not even using an index (inverted index)

wow, so basically it took 300 ms to group 37M docs so with an inverted index it should be faster

Not for your query as there is no filtering 

Here's a sample config for star-tree index : <https://docs.pinot.apache.org/indexing#index-generation-configuration>

<@UDRJ7G85T> ah really? group by is not using index? so it is pointless to create an inverted index (<https://docs.pinot.apache.org/indexing#inverted-index-only-available-with-dictionary-encoded-indexes>) for the column i need to group