I currently have following granularitySpec and I was wondering if choosing different segmentGranularity/queryGranularity might have a cost impact? e.g., if choosing larger queryGranularity like

fifteen_minute

or

hour

could reduce cost.

"granularitySpec": {
        "type": "uniform",
        "segmentGranularity": "DAY",
        "queryGranularity": "MINUTE",
        "rollup": true,
        "intervals": []
}

Hi Julian, Rollup is essentially a GroupBy on __time and the dimension fields, so the coarser the granularity of __time the more likely rows will group together, resulting in lower row count and therefore smaller disk footprint and faster query processing on the data. Is that along the lines of the info you were looking for?

All right, yeah thats what I was looking for. The larger the granularity is the lower row count will be and the less storage will be used right?