RE: Actual row size of an index
All the tests we have done have shown that OLTP benefits from the smallest pagesize because of the contention on pages. The only exception to this is when you can move to the next larger pagesize and reduce the total number of pages in a table by a large percentage. This allows more rows to fit in the bufferpool. Example a table contain 100,000 rows with a 1,300 rowsize pagesize # of page to store all rows space required to= store data 2KB 100,000 200,000 KB 4KB 33,334 133, 336 KB 6KB 25,000 150,000 KB John = "Robert = Roussey(IT)" = <Robert.Roussey@S = To piritAir.com> ids@iiug.org = Sent by: = cc ids-bounces@iiug. = org Subj= ect RE: Actual row size of an index = [10997] = 01/18/2008 10:19 = AM = = = Please respond to = ids@iiug.org = = = That being said, would there be an advantage - if you have a large rowsize - to sizing large pages to get close to 255 rows on them for OLTP? Art? JM3? Bob Roussey Unix / Informix Administration Spirit Airlines Robert.Roussey@SpiritAir.com -----Original Message----- From: ids-bounces@iiug.org [mailto:ids-bounces@iiug.org] On Behalf Of kernoal.stephens@autozone.com Sent: Friday, January 18, 2008 12:59 PM To: ids@iiug.org Subject: Re: Actual row size of an index [10996] That slide was wrong. I have tested this and you can fit a large number of index rows on a 16= K page. Data rows are still 255! Huge performance increase for OLTP queries. FYI Kernoal "Art S. Kagel (Oninit LLC)" <art@oninit.com> Sent by: ids-bounces@iiug.org 01/18/2008 11:50 AM Please respond to ids@iiug.org To ids@iiug.org cc Subject Re: Actual row size of an index [10995] MICHAEL FAVOLE wrote: > Thanks for the reply Art. We tested with a 16k size for our critical indexes > and had positive results. > > I am a bit confused about one thing though. In a chat with the labs > presentation (IBM Informix Dynamix Server 2005-11-02 slide 10), Jonathan > Leffler indicates that there is "Still a maximum of 255 keys per inde= x page". > So this is probably not correct, agree? > Dunno now. That's not my understanding, but I've been wrong before - mostly when I contradict Jonathan ;-) Jonathan? Comments? Art S. Kagel Oninit > Thanks again, > Mike > > > =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D Please access the attached hyperlink for an important electronic communications disclaimer: http://www.oninit.com/home/disclaimer.php =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D ***********************************************************************= * ******* Forum Note: Use "Reply" to post a response in the discussion forum. See you at the IIUG Informix 2008 Conference The Power Conference for Informix Professionals April 27 - 30, 2008 Marriott Overland Park (Kansas City), Kansas http://www.iiug.org/conf Registration Now Open!! ***********************************************************************= * ******* Forum Note: Use "Reply" to post a response in the discussion forum. See you at the IIUG Informix 2008 Conference The Power Conference for Informix Professionals April 27 - 30, 2008 Marriott Overland Park (Kansas City), Kansas http://www.iiug.org/conf Registration Now Open!! ***********************************************************************= ******** Forum Note: Use "Reply" to post a response in the discussion forum. See you at the IIUG Informix 2008 Conference The Power Conference for Informix Professionals April 27 - 30, 2008 Marriott Overland Park (Kansas City), Kansas http://www.iiug.org/conf Registration Now Open!! =