Database in Trading part 2

6 日前

TLDR; ArticDB looks to be the best current option.
CSV - Simplest, slowest, largest and no data types can lead to type errors when loading.
SQL Database - typically transactional row store (OLTP), slow for analysis (OLAP), difficult to shard and parallelize workloads over clusters.


While MongoDB proved the fastest to read and write end-of-day OHLC (Open, High, Low, Close) data — the SQL solutions were 1.5 × to 3 × slower — in terms of costs MongoDB was definitely the most expensive due to its commercial licensing costs.