Hi Guys,
I am dealing with a 3rd party Progress App for which we only have r-code.
The database that comes with this app has the following characteristics:
So for obvious reasons and the tables (800 in total) should be re-organized into different data areas.
I have a couple of questions:
Your input would be very much appreciated for this.
This will not affect CRC, so you won't need a recompile (although testing is still a very good idea!)
You can get all fancy with the distributions, and you may get some benefits, but a 62GB database isn't that big. You will get most benefit from splitting tables, indexes and LOBs into separate areas.
If you do want to do some separation, then split off your busiest tables into one area, put the indexes for those tables into another area, and then stick your remaining tables into an area with the indexes again separate.
This will not affect CRC, so you won't need a recompile (although testing is still a very good idea!)
You can get all fancy with the distributions, and you may get some benefits, but a 62GB database isn't that big. You will get most benefit from splitting tables, indexes and LOBs into separate areas.
If you do want to do some separation, then split off your busiest tables into one area, put the indexes for those tables into another area, and then stick your remaining tables into an area with the indexes again separate.