Re: Static or Dynamic rows -- which is faster?

2002-02-12 Thread Heikki Tuuri
Hi! -Original Message- From: David Felio [EMAIL PROTECTED] To: Heikki Tuuri [EMAIL PROTECTED] Date: Tuesday, February 12, 2002 12:47 AM Subject: Re: Static or Dynamic rows -- which is faster? On Monday, February 11, 2002, at 12:12 PM, Heikki Tuuri wrote: for InnoDB 'dynamic rows

Static or Dynamic rows -- which is faster?

2002-02-12 Thread Eric Mayers
I have a large data set (15 mil rows) consisting of a datetime column and a char(255) column. I seem to recall seeing something about performance benefits for using static length rows with MyISAM tables, and I've heard some argument for using dynamic length rows (e.g., use varchar rather than

Re: Static or Dynamic rows -- which is faster?

2002-02-12 Thread Heikki Tuuri
Eric, for InnoDB 'dynamic rows', that is, rows where you define char columns as VARCHAR, are faster because tables and indexes fit in smaller space. Only in some rare cases where you want to avoid fragmentation caused by updates which change a column length, a fixed-length CHAR(...) column can

Static or Dynamic rows -- which is faster?

2002-02-12 Thread David Felio
Whoops, originally sent this to just Heikki. On Monday, February 11, 2002, at 12:12 PM, Heikki Tuuri wrote: for InnoDB 'dynamic rows', that is, rows where you define char columns as VARCHAR, are faster because tables and indexes fit in smaller space. Is there reasoning specific to InnoDB?

Re: Static or Dynamic rows -- which is faster?

2002-02-11 Thread Jeremy Zawodny
On Mon, Feb 11, 2002 at 04:48:10PM -0600, David Felio wrote: Whoops, originally sent this to just Heikki. On Monday, February 11, 2002, at 12:12 PM, Heikki Tuuri wrote: for InnoDB 'dynamic rows', that is, rows where you define char columns as VARCHAR, are faster because tables and indexes