Typical "canned" answer is try to partition by a numeric. We partition by
a date when we can.
I'm torn between the order of one and two, but that is more personal
preference. I generally try to partition by the fields that make joins run
best and the ones that make maintainence better if I c
Bill - Since nobody has replied yet, I'll toss in a couple of ideas. What is
your motivation for partitioning?
Performance? That is what I get from your posting. I think you have good
ideas. I'll provide one more that got us a good performance boost. Some
queries were often comparing "this mont
Bill,
I'm impressed! Most of us follow something similar to the rankings you
mentioned - but a very few actually spell it out and put it for the data
architects. Now that you have done most of the work, I have a few comments.
First, you don't have to worry about legal issues for testing these out