Re: acceptable number of extents?

MDevlin inquires: >>our database has over two hundred tables, 9 of them have more than one >hundred >>extents and another 35 have over 10. how signifigantly could this impact >>performance? To which Clem replies: >you may now be fairly castigated for not looking in TFM "Performance >Guide", where it can be found in the discussion on extents in >chapter 4. Let me take this opportunity to brag about Informix's >documentation, which is *well* worth the time to read. It's the >best in the business. Prior to V7 extent fragmentation was much more important to stay on top of. These algorithms are more efficient now. So much so that if you are looking into a performance issue I would start elsewhere. I have consolidated tables in excess of 180 extents with no measured performance difference. You may run out of the ability to add extents. One of Informix's hairiest formulas covers this. My mind went numb trying to figure this formula once and I just went ahead and did the consolidation! OTOH if you have little else to do run some timings, consolidate, time again and post your results here. As is the case with many other highly technical questions a definitive answer is not provided in TF Performance Guide (V7.2 home copy). I have five or six different Online tuning guides and on some questions I can find five or six completely different answers and on other questions, none. Having read through MANY manufactures doc sets I do appreciate Informix's very good work. The documentation falls apart at the high end though. Enjoy, Bill