4 ms·
QuestDB – OpenSource Time Series Database
I would like to plug new time series database, which I'm writing: QuestDB (https://github.com/bluestreak01/questdb)
This project was born out of the realisation that data can be handled much faster than existing databases would lead you to believe, and a general fascination in how fast hardware can do these things provided you ask it the right things to do.
This brief history aside, the aim of QuestDB is to provide a full fat SQL dialect for querying data,
including time series, simple, fast and convenient tools to import, export and otherwise integrate data into your existing workflow.
To that effect I have done quite a lot already and would like to invite you to check out list of features on my github page.
I would be delighted if you could look at the source code and ask whatever questions you may have and of course leave comments however harsh they may be.
I'm hear to learn.
Thanks,
Vlad
- dozzie 10y agoWhy every new timeseries database needs to implement SQL dialect, which is awful when it comes to time-related data? Why nobody looks for other languages and data processing models anymore? There were so many research papers about that!
- bluestreak 10y agoIn this case rationale for SQL was relatively mild learning curve. Surely it is a good thing? But you raised a good point. In QuestDB SQL is implemented into syntatic sugar over java-based streaming model. Writing support for another language wouldn't be that hard provided there is consensus on what suitable query language is?