Please start any new threads on our new
site at https://forums.sqlteam.com. We've got lots of great SQL Server
experts to answer whatever question you can come up with.
| Author |
Topic |
|
Pradip
Starting Member
32 Posts |
 Posted - 2009-11-25 : 07:50:59
|
| I have transaction table of bank account and having one column accountID 10 digit char. I want to have Index on it, I think clustered index is better option, can anyone tell me why not non-clustered?pradipjain |
|
|
Transact Charlie
Master Smack Fu Yak Hacker
3451 Posts |
Posted - 2009-11-25 : 08:52:08
|
| It might be hard to say for sure (obviously without knowing the table structure) A few potential considerations though:The first: You can only have 1 clustered index. Therefore you want to get the most bang for your buck. Is that accountID the column that you would search on the most in the tabke? Is it the one generally joined to and referenced in where clauses. It probably is I imagineThe Second: Insert performance. Best insert performance is with a strictly increasing clustered index (new rows are always greater for the clustered index column that old rows). Is that the case here? Probably not as it is a table of transactions.Would be a hard call to make. Is there not a clustered index on the table at the moment? the table is a heap right now?it might come down to how often the table is read from or written to. If mostly writes then a strictly increasing CI might be better with whatever covering non clustered indexes implemented on the other columns.Charlie===============================================================Msg 3903, Level 16, State 1, Line 1736The ROLLBACK TRANSACTION request has no corresponding BEGIN TRANSACTION |
 |
|
|
|
|
|
Cluster or Non Cluster