We have a long running query that is working on a main table that has ~ 1 Billion rows.
The query does a group by and creates aggregates along with calculations and also CONVERT statements...it has 4 joins with the other 3 tables having ~ 333 Million rows. Indexes seem ok, execution plan looks ok.
Box is a 64 bit 2 proc with 4 cores each and 24 GB or RAM. The large table is ~ 75 GB w/out indexes.
What we are seeing is heavy use of tempdb both reading and writing, tempdb has grown to .4 TB in size.
Question: does this as you would expect in terms of resource utilization and time (1 day running so far)? Any ideas or advice on what to look at in terms of efficiency and also how to tell how far along it is in the the process of crunching thru?
thank you.
Rick