eXtremeDB Financial Edition – Sequences Demo
The ‘Sequences Demo’ program shows eXtremeDB Financial Edition’s performance attained using its ‘sequence’ data type for managing time series data – in this case, trades and quotes (TAQ or market data).
The demo emulates a tick stream for a single symbol over the course of a day, compressed to a short time period (a very active stock might have 10 million ticks in a 6.5 hour trading day). It loads the tick data into the eXtremeDB Financial Edition database at maximum speed via a single thread; multiple reader threads can run in parallel.
Reader threads simultaneously perform several common analytical queries supported by eXtremeDB Financial Edition – including Sum, Covariance and Correlation – in looping fashion against the tick stream. The demo measures performance, displaying metrics including ticks processed per second and calculations per second.
Download the Sequences Demo.
Get the readme file with more details about the demo.
Sequence Data Type
eXtremeDB Financial Edition’s ‘sequence’ data type implements columnar (as opposed to row-based) handling for a given data element, as part of a database design.
Traditional database management systems (DBMSs) bring rows of data into L1/L2 cache for processing. But financial data – such as ticks, trades and quotes – are better handled by a column-based layout that maximizes efficiency in fetching needed information. The result is faster performance: the database system benefits from L1/L2 cache speed and avoids fetching data via the much slower front-side bus.
eXtremeDB Financial Edition’s ‘sequence’ data type implements columnar handling
for specified data, resulting in greater efficiency and faster performance.