On 7 Oct 2011, at 5:30pm, Mohit Sindhwani wrote: > I have been trying to see how we can make one of our databases more space > efficient. I am now looking at seeing if we have the right indexes and if > there is a way to save space by removing indexes. > > Is there some way to make use of the fact that the data can be sorted by a > specific column in the desired order before it is imported into the table, or > do we need to create an index on that column anyway?
You need an index on that column and that column only if something is going to use it. > Eliminating the index could save us space. One of the things in our case is > that the data is in a read only database (and further compressed/ encrypted > by CEROD). The most important thing to note is that having an index on each column is completely different to having indexes which speed up your SELECT and UPDATE commands. Many people starting out with SQL think that indexing each column is a useful thing to do. Instead you should be looking at each SELECT and UPDATE command and trying to work out indexes which would be useful for those commands. Simon. _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users