DEV Community

[Comment from a deleted post]
Collapse
 
rhymes profile image
rhymes

Dealing with massive tables, hundreds of millions or billions, when adding an index might actually slow down inserts πŸ˜‚

Collapse
 
helenanders26 profile image
Helen Anderson

I'm dealing with this right now, each analyst has an opinion on indexing and if I add all their suggestions the table will never update

Collapse
 
aarone4 profile image
Aaron Reese

If you have that much data you should hopefully be on an Enterprise level product and partitioning should be a viable option. If you are on mssql, get onto sqlservercentral.com forums and reach out to gilamonster or grant Finchley. They should be able to help with strategy

 
helenanders26 profile image
Helen Anderson

We have 70+ analysts on AWS Aurora (Postgres). The clustered indexes are applied by my team (BI), after that it’s up to the analysts to agree how they would like extra indexes to be built to support what they are doing