Deciding which repository to use¶
Valo has two repositories:
This page helps you decide which one you should use.
Will you need to perform historical analysis on the stream’s data?
If not - i.e. you only need to perform real time analysis - then you don’t need to use a repository at all.
Is your data hierarchical in nature (i.e. does it have a nested JSON structure)?
If so you need to use the SSR, because the TSR doesn’t support hierarchical data.
Does your data have a flexible or unknown structure (i.e. would it be hard to produce a schema for your data)?
Alternately do you want to begin storing data without specifying a schema?
In either case you need to use the SSR, because the TSR requires that streams have a schema and that payloads match the schema exactly (bearing in mind schema fields can be marked as optional).
Do you need to perform mostly text based analysis, or do you need to use free text search?
If so you should use the SSR.
Do you need to perform mostly numeric analysis, or do you need to be able to search by the shape of a graph?
Is your data represented as a time series? Do you mainly want to query it in the context of time?
If so you should use the TSR.
Do you generally only select a handful or the available fields present in your data?
If so you should use the TSR as it is based on column store technology so is likely to be faster.