pikers/piker

tsdb multi-backend support and testing!

Opened this issue · 0 comments

As per the slew of issues pertaining to time-series-db subsystems in piker seen here:
tsdb time series db stuff


Hi-level ToDo:

  • breaking out the middle-ware history managment layer into it's own subpkg as per #315
  • trying out at least one other tsdb (probably arctic) and formalizing our OHLCV history apis as per the research options listed in #90
  • actually documenting how to launch the db with --tsdb 😂 as per #313
  • generally speaking, much better management tools for manually pulling data from providers and populating a db instance either with cli or context-menu options in our charting app!
  • real-time ingest of tick and signals (from our fsp layer) as per, for eg. #314
  • a tick history to 1s OHLCV downsampler system that can be used cross-provider since most don't provide this low of a sample rate (think only one is ib?)

...

maybe more?


testing ToDo:

  • db client connection, write, read and delete a time series loaded from either an offline sample or some provider we can always rely on (like kraken maybe?)
  • basic tests for replication for those dbs which support it