Now I'm choosing my indexes, so I have another question: I have a bigint column with 90% different values, the duplicate values are rare, but the column is not very...

If I make log backups ...the log space will be reusable, but in order to use the log information I will have to restore the log backups I think...am I...

So shrinking the log file has no side effects? Especially on performance

Because I want to shrink it but I'm affraid of unknown side effects.

I know that shrinking the database has...

Ok man...message received.I will search for other candidates on clustered index, and if I don't find I will put them on guids.That's the situation.Thank you...things are more clear now:-)

So it's better to leave them like this, non-clustered, than put a clustered index on pk?

I repeat, they are mainly used for joins, so a clustered index on other columns...

The guids in the pk are generally used for joins...so I have to change the index in a clustered one...

I just don't know why they are created this way....I don't...

ok...

But I observe smth on the db...all the big tables don't have a clustered index on PK. Is this a technique, or something? Because I don't think this is a...

I found the problem, I think

The Pk-indexes are non-clustered,and that's why there are heaps. I don't know if this is ok, or if the person that was here before me...

So

the result from sys.dm_db_index_physical_stats :

4824148798901HEAPIN_ROW_DATA1093,36099585062242267,446902654867261683NULLNULLNULLNULLNULLNULLNULLNULL

4824148798901HEAPLOB_DATA100NULLNULL463NULLNULLNULLNULLNULLNULLNULLNULL

48241487989111NONCLUSTERED INDEXIN_ROW_DATA203,404255319148942210,6818181818182235NULLNULLNULLNULLNULLNULLNULLNULL

48241487989251NONCLUSTERED INDEXIN_ROW_DATA201,271186440677971219,6666666666667236NULLNULLNULLNULLNULLNULLNULLNULL

The result from you query :

2414879890NULL

24148798911Pk_Mt_Anag

24148798925ix_pratica_id

Yes, is normal for indexes, clustered and non-clustered, but if I have a heap and a clustered index for the same object_id.This shouldn't be normal, if I folow what you...

let me ask in this way: if I query sys.dm_db_index_physical_stats and get the same object_id for a heap and an index is an abnormal situation?

Because I will paste some results...

Man ...I agree with you, but when I run the query against sys.dm_db_index_physical_stats I receive an object_id, and object_name(object_id) returns a table who has for instance 2 indexes,1 clustered, and...

First, as you can see in myreply with the code, I rebuild all the indexes. For some, as I said before, I don't know why but they still have a...

nice discution....

Now i'm "optimizing" my optimization.:-)

That's why I have some more questions 😀

I found that the indexes with more than 1000 pages were rebuild, and I have some others...

17 pages, but I have some with 2 pages or 34 that "respond" the same way to my script

And no, is not on production, I've restored the database on...