3 replies
by srbc
There might be many different reasons. Your SQL server configuration, the design of the database it self, indexes and primary keys (usualy when you import table from another platform, primary key and index setting might not be imported with it therefore resulting a much slower data retrieval), nested triggers and ect.
So there is no single answer for your question. I will recommend reviewing your database from A to Z completely and you might be able to find the botle necks.
Is the SQL server on it's own machine? How big is it?
It sounds like the front end didn't get migrated yet. The biggest bang would be from converting embedded SQL in the front end into stored procedures. This cuts down on network traffic and makes the server do the work, rather than the same front end machine, or user machine.
If you have re-written the application in windows then you will take some performance hit due to the nature of a graphical environment.
Possible things to look at are
1) indices on tables
2) how much data your SQL is sending over the network. (using stored procedures can help limit this)
3) The type of locking you use on tables.
4) whether you use client side or server side recordsets.
One way I find useful when looking for bottlenecks is capturing the sql string which is taking a long time and executing it directly via query analyser. You can see how long it takes and also look at the execution plan which may give you hints on how to improve performace if the SQL is indeed the bottleneck
This discussion is closed to new replies.
Trending Software posts
Sort by Trending