sql - Non-Clustered Index on all the columns -
if allowed create 249 non-clustered indexes on table mean safe have non-clustered on every column? effect of that? or have choose columns create non-clustered index.
is this:
create nonclustered index ix_test(col1, col2, col3)
different this: ?
create nonclustered index ix_test(col1) create nonclustered index ix_test2(col2) create nonclustered index ix_test3(col3)
indexes reads , bad writes.
when reading data indexes can find data quicker , make reads quick.
on other hand
when writing data (update/delete/insert) sql server have write data twice once on sql server pages actual data stored , once indexes. result in slower writes.
you have find middle ground reads not suffer not enough helping indexes , writes not suffer many indexes.
the above statement true oltp applications, data changing quite if olap/data warehouse data hardly changes if ever, can add many indexes reads require perform better. hope helps.
Comments
Post a Comment