We have a large BI data warehouse. The database is updated every night with the previous day worth of activity, all week to date, month to date, and year to date numbers are reaggregated, etc etc. Once this data is processed in overnight, its baked. Nothing else updates it for the rest of the day - it is only ever read as people request reports.
On particularly busy days, especially when several different people in separate departments are requesting variations of the same data, we'll see some slowness. While I believe this is mostly unavoidable and is simply the result of pages being read from disk, I was wondering if a situation like this would be a viable candidate for adding NOLOCK to some of these tables in the bigger queries. I figure it may help speed along some of the queries where I'm seeing many people hitting the same data at...