Preferred Member from GB
joined:July 25, 2005
I'm developing a multi-user toolset. We're talking about 5 data-intensive tools and predicted 1,000 users.
Currently each tool has a separate table and whenever I run a tool, it dumps up to 20,000 rows into a table. It's ok now that I'm the only person using the toolset. What happens when other users join? Tens of millions of rows!
Some tools are interlinked and querying/comparing several tables.
Somebody suggested I should automatically create a set of personal tables for each user who signs up, but then it's 5,000 tables or potentially more.
From the perspective of server resources and page load times, what would you choose? Keep five massive tables or deal with thousands of small tables?