rows v estimated rows | SQL Server Performance Forums

SQL Server Performance Forum – Threads Archive

rows v estimated rows

I’m troubleshooting a slow running query . I’ve saved the execution plan and noticed that one of the
estimated rows = 378
estimated executions = 1
Rows = 15469
Physical Operator = Hash Match
Logical Operator = Aggregate
The query is accessing a remote db , on a column that has no index (although there is a clustered index . ) Essentially, this is part of an INNER JOIN. The WHERE part of the statement , is looking for a value on the local db not the remote I’m looking for ways to think through this problem – for example, as the column doesn’t have an index , does that mean a non clustered index would help , why? Thanks in advance for any advice

If that is the only column in your WHERE clause and it is not a part of the primary key, you may add a non clustered index on it and see the performance. However, it also depends upon the columns specified in the select list. I would suggest you to go through some of the articles on this website relating to query tuning and indexes. Regards, Chetan
Is the query joining with a table on the remote db? or are all tables in the query on the remote db. Can you post it? Thanks! Michael B
Sr. DBA "The fear of the Lord is the beginning of knowledge,
but fools despise wisdom and instruction." Proverbs 1:7