Re: Overflow Distributions in Online 7.20

mc wrote: > > I am having a performance problem in Online 7.20. When I look at the > distributions on the table using the "dbschema -hd" command , I notice that > the leading column of the index which I am using only shows 3 distributions > "bins" but shows 200 in an OVERFLOW heading. What causes this overflow > info on the distribution report and is it related to slow performance on > using the index ? An overflow 'bin' is created for any key value whose occurrance count is more than a certain percentage of the total count for the keys in the bin to which it would otherwise have been assigned. I asked an intense series of questions on this very subject a few months ago. I saw something similar and wanted to know if I should adjust the UPDATE STATISTICS parameters to widen the bins so that fewer keys were in overflow bins or narrow them so that more were. The answer is that overflows are a "GOOD THING" they let the optimizer know exactly how many of each of the most populous key values there are. This helps the optimizer to zero in on those values when needed and also keeps the estimates for less populous keys from being skewed by these keys that are essentially outliers. Art S. Kagel