We would need to have something more to go on. Like table structure, query, execution plan, indexes. And index can help for a seek (predicate, search condition) for the outer table in the plan, for the other tables, it is about helping with the join operation. So in the end it boils down to your queries, search conditions, indexes, etc.
UniqueIdentifier - Performance Issues Though doing a Index Seek
Dear Experts,
I have two Tables with 800 Million Records in each where the Two Tables which I have Uniqueidentifier Column in the both as Primary Keys and one is references to other table.
While the Select is done with other tables including these tables as Joins, I have severe performance issues though the Execution plans shows the Index Seek Operation happening on these tables but the query cost on these two tables shows 70 % for these two tables alone. Also to mention these tables are partitioned with the timestamp without a purge policy as the data is important.
Indexes:
Primary Key (Clustered Index) Uniqueidentifier
Non-Clustered Index (Partitioned Column)
Any help on improving the queries performance could be greatly appreciated
Thanks
Momen
2 answers
Sort by: Most helpful
-
-
Tom Phillips 17,731 Reputation points
2020-08-21T12:11:27.77+00:00 As Tibor said, we would need to see your query and your execution plan to give any information. There is nothing inherent in a uniqueid which would cause a performance issue on queries, and you indicate it is already doing an index seek. Having 70% being taking on a huge table is not an indication of a problem, just where it is spending most of the time.