RE: Problem with statistics

The batch program does updates on one of the tables mentioned. This table contains =/- 1.600.000 records and the program changes approximately 2000 - 3000 records (including index values). What does your script ? Does it perform update statistics high or low ? If I use only selects that follow index path, isn't a simple update statistics enough and why not ? -----Original Message----- From: Obnoxio The Clown [mailto:obnoxio@hotmail.com] Sent: dinsdag 28 maart 2000 12:35 To: doris.willaert@suzuki.be; informix-list@iiug.org Cc: obnoxio@hotmail.com Subject: RE: Problem with statistics From: Willaert Doris <doris.willaert@suzuki.be> > >1) Why so often ? Normally we did the update statistics once at night. But >we saw that after running a batch program early in the morning, the >optimiser chose the wrong path. What does the batch program do? >2) Yes, this is what we see. We also saw that there was a program running >that apparently does "bad written" selects. After this program was killed >and after we had done an update statistics, the select was ok again. We run >a simple update statistics. <PLUG> Have you tried my update stats shell script from http://www.iiug.org? It might make a better, more permanent solution. </PLUG> ______________________________________________________ Get Your Private, Free Email at http://www.hotmail.com