DataGeek & the 17 indexes...
DataGeek is digging around today and he's come across a common issue, a table with way too many indexes.
Seen this before?
Is there a "golden rule" for too many indexes on a single table? No, there really aren't any good "golden rules" but it's somewhat safe to say that a "guideline" would be that if the table is mostly read only (mostly SELECT statements, not many INSERT/UPDATE/DELETE statements) then you can index the table as you need. If the table is "more normal" and has a normal mix of SELECT and DML (INSERT/UPDATE/DELETE statements) then you need to be considerate of the performance issues your DML statements could have by over-indexing.
The next step for DataGeek would be to determine what SQL statements are being run on this table and figuring out which indexes are helping and which can be removed.
Until next time...Rich
Related Post: See how DataGeek tries to talk tough here...
Seen this before?
Is there a "golden rule" for too many indexes on a single table? No, there really aren't any good "golden rules" but it's somewhat safe to say that a "guideline" would be that if the table is mostly read only (mostly SELECT statements, not many INSERT/UPDATE/DELETE statements) then you can index the table as you need. If the table is "more normal" and has a normal mix of SELECT and DML (INSERT/UPDATE/DELETE statements) then you need to be considerate of the performance issues your DML statements could have by over-indexing.
The next step for DataGeek would be to determine what SQL statements are being run on this table and figuring out which indexes are helping and which can be removed.
Until next time...Rich
Related Post: See how DataGeek tries to talk tough here...
Comments