mk::select using -globnc is all of a sudden very slow. My db contains 282,000 records and works great with mk::select returning in less than 1 second. I added an additional 3,000 rows and now it takes 30 secs for mk::select to respond. My question is how do I go about figuring out what the problem is? I have done a lot of testing and feel the problem is not related to size or the content of the new records.
If anyone can give me a hint on the best way to diagnose such a problem I would be greatful. I can make the databases (before and after) available if anyone wants to look at them.
- 2007-10-10 Allen Seifert
Created
- 2007-10-15 EAS
The databases magically returned to normal so no crisis for now. I would however like to get hints on how to diagnose a slowness problem as I described. BTW, this is not the first time a slowness that magically went away has happened. And it's not a machine or network problem because it happens in different locations on the same db file.
- 2007-11-18 jcw
(Changed: stat)
- 2007-11-19 Allen Seifert
Hi JC - I see you recently changed the Status of this item to "review". The meaning of Status values I see in the FFF seem obvious but I wonder, is there any documentation on exactly what they mean?
I heard from my customer today. She asked if we heard anything about this slowness problem. They are getting ready to add another batch of entries into their database which caused the problem last time around.
Thanks in advance for any tips you can give in helping to diagnose the problem.
PS. I will be receiving soon the current version of their database which is after the slowness problem "magically" went away. I hope it will help me figure this out.