You have a nonpartitioned table that has a single dimension. The table is named dim.Products.Projections.
The table is queried frequently by several line-of-business applications. The data is updated frequently throughout the day by two processes.
Users report that when they query data from dim.Products.Projections, the responses are slower than expected. The issue occurs when a large number of rows are being updated.
You need to prevent the updates from slowing down the queries.
What should you do?
A. Use the NOLOCK option.
B. Execute the DBCC UPDATEUSAGE statement.
C. Use the max worker threads option.
D. Use a table-valued parameter.
E. Set SET ALLOW_SNAPSHOT_ISOLATION to ON.
I would go for E. Another option would be using nolock but I’d rather not use that because of the dirty reads scenario. If tempdb performance allows I would prefer E.