I joined a company that did twice yearly price updates. They had 10.4M price points across 60K skus and seven marketplaces. The update had to be run in real time (i.e. the night of the price change) and took 5 hours and has to be done by the IT team.
I made it so that the users could update the prices in advance and the application on the day took 9 minutes.
Several things. Firstly introduced some efftive from/to fields so that future records could be entered. Second removed a whole bunch of cursors. Third new indexes. Fourth lots and lots of temp tables. Fifth proper table joins so we could make use of statistics. Sixth. Introduced some config and aggregator tables.
Was an interesting project.
For further actions, you may consider blocking this person and/or reporting abuse
We're a place where coders share, stay up-to-date and grow their careers.
I joined a company that did twice yearly price updates. They had 10.4M price points across 60K skus and seven marketplaces. The update had to be run in real time (i.e. the night of the price change) and took 5 hours and has to be done by the IT team.
I made it so that the users could update the prices in advance and the application on the day took 9 minutes.
Did you optimize query or something ?
Several things. Firstly introduced some efftive from/to fields so that future records could be entered. Second removed a whole bunch of cursors. Third new indexes. Fourth lots and lots of temp tables. Fifth proper table joins so we could make use of statistics. Sixth. Introduced some config and aggregator tables.
Was an interesting project.