Re: About UPDATE triggers...
Posted in 1997
In article <5msa0q$r2h@cssun.mathcs.emory.edu>, MarkDenham@aol.com wrote: >As far as informix goes I would bet that it is faster to just do the update >than to worry about checking the value first and then allowing the update. > I had done a study of log records some time back and all indications were that Informix automatically ignores columns which do not need to be updated (because they aren't changing). The test clearly indicated that 1) Log records were only created for rows which changed (not those which qualified in the WHERE clause) 2) The size of the log record was related to the size of the change and not to the row size. >The only time you might want to avoid unnecessary updates to columns is when >the column is indexed. The same applied to the index. Even if a row changed, its index rows would only be logged if the actual columns which were part of the index changed. > >Mark >MarkDenham@aol.com (home) >Denham.M@amstr.com (work) ----------------------- Rudy Fernandes GIC, Kuwait OL 7.20UC4, 4GL 6.04UC1 -----------------------