Is there a feature available where time series based data could be aggregated for all fields of an index and and stored as another aggregated document in another index. To elaborate, say index1 has time series data and it is collected every 5 seconds for numeric fields nfield1, nfield2 and nfield3 and some text fields (which can be used for grouping the documents aggregation on these numeric fields).
In summarized index I want to keep data at much coarse granularity over time. Say these numeric fields are aggregated (could be average value by each hour or day or week). Thus summarized index contains one document (for the grouping used) instead of 1260 raw documents stored for one hour period in original time series based index or 1260*24 documents averaged for one day creating one document for quick access and trending from summarized/aggregated index instead of computing on fly for different aggregation periods.