I found a great article on wait locks here on SQL ServerCentral, but was wondering if someone could expound on what the author was saying regarding the below:
"You can adjust this level so that SQL server will not consider lower costing queries for parallelism, and by profiling the queries and procs your can find the sweet spot for your server."
Full Article: http://www.sqlservercentral.com/articles/Administration/wait_types/856/
My question is, what determines a low costing query vs high costing query as it relates to parallelism? Is it size of the tables involved or perhaps gnarly joins?
If I have a very simple query like the below, but both tables involved have...say...150 million records, would high parallelism or low parallesim be optimal for this query?
FROM Table1 t1
INNER JOIN Table2 t2 on t1.Col1 = t2.Col2