How to store/source/present long-term time series (which component to use at which place)

Greetings friendly people :vulcan_salute:

I am trying to piece things together and do not know how to put things together.
I HAVE a database with kinda time-series data, some statistics that I measure every week since 2 years and I want to visualize the long-term evolution of those statistics. (many different time series with ids)

The data itself is in a small database (currently oracle). I have a web application which I can extend at will, e.g. to write a URL that provides the required data as JSON or CSV. Producing new data is done “somewhere else” and should not be changed.

Now I am looking for suggestions how to piece things together, any input is welcome!

  • Prometheus? As far as I have seen, this has its own storage but it designed for short-term storage. It will not be interested in numbers I created 2 years ago.
    I have seen there is a “push gateway” to push older data into it but not sure if this is the right thing for my puzzle

  • Infinity data source connecting to a custom URL that I will write providing JSON?
    I would need to know the DETAILS, mostly caching and querying behavior. The simple version is to provide a JSON with ALL data but as I understand, each time I repaint a diagram, that URL will be called and the DB queried. I would expect the datasource to cache some data and only refresh new items?
    Also I would like to know how to make infinity smart e.g. to provide the requested time range to the backend (so that the JSON does not deliver 2 years of data when the chart is about last week). I looked at the documentation but it was vague, telling me about “backend parser option in queries” and I know nothing about that field.

  • Oracle data source? Uhm… no, thanks. budget and I dont want to monitor oracle, I want to retreive a few numbers. And want to avoid direct DB connection.

  • anything else that I missed so far?

Thanks in advance!