Unanswered: Update 40+ million row table..speeeed.
I have a big table (40+million rows) with 135 columns. On a daily basis I need to update this table setting the content of only one of the columns. Problem is that updating this entire table will be a very time consuming process, plus depending on how we build the query it can take forever to run.
Can anyone give me performance tips that will acomplish what we need in the fastest time possible? I read somewhere that /*+ parallel(x,10) */ will help but not sure..
Parallel could help, if you have a number of processors to spread the query over - and if parallel query is enabled on your database. Your query will hog the server while it is running, but should run quicker.
On the other hand, is updating every row of a 40M row table every day necessary? What is this update? Maybe there's an alternative solution.