Hi
I use zeoslib 6.x.x with Mysql, in program that query large Mysql data base locally connected ( I delver Mysql with this program so there is no server for end user).
My problem the open and close take much time, when this database reach 5000 record records has many 255 car. Varchar and some graphics stored in this database.
So please if there is solution to optimise this, I have tried 6.5.1,6.6.2 and I have the same issue
Thanks and excuse my English
Mysql Open and close query very slow
Moderators: gto, cipto_kh, EgonHugeist
-
mdaems
- Zeos Project Manager
- Posts: 2766
- Joined: 20.09.2005, 15:28
- Location: Brussels, Belgium
- Contact:
Did you try this using other connectivity components or even using libmysql directly? My idea is this is mainly due to the workings of libmyqsl.
Some ideas to help your application forward.
- If you need to search the records regularly you shouldn't use a TZReadOnlyQuery
- If you are using a grid to show the data you might want to use the patch in http://zeos.firmos.at/viewtopic.php?t=1857.
- Even if 5000 records isn't a big dataset : avoid fetching every row and every big field. You rarely needs 5000 large records at once.
Mark
Some ideas to help your application forward.
- If you need to search the records regularly you shouldn't use a TZReadOnlyQuery
- If you are using a grid to show the data you might want to use the patch in http://zeos.firmos.at/viewtopic.php?t=1857.
- Even if 5000 records isn't a big dataset : avoid fetching every row and every big field. You rarely needs 5000 large records at once.
Mark
Last edited by mdaems on 13.08.2008, 10:27, edited 1 time in total.
First Thanks for replay.
I have change all TzReadOnlyQuery to TzQuery a few days agoo, But seme result.
I am not the designer of Database,so I can't change any things on it, also there is Unicode (arabic) field in main Table and others may this affect the speed, I meen Unicode field.
The database is composed from 33 tables linked to main table , and all research is done in main table and then linked to other table with masterdetail to main table.
I have use Corelab MyDac and the speed is correct, but I don't like the core lab ( flowwork) and is very expencive to bay this component for only this issue they don't gave more function than zeoslib.
Eny more thanks for you idias.
I have change all TzReadOnlyQuery to TzQuery a few days agoo, But seme result.
I am not the designer of Database,so I can't change any things on it, also there is Unicode (arabic) field in main Table and others may this affect the speed, I meen Unicode field.
The database is composed from 33 tables linked to main table , and all research is done in main table and then linked to other table with masterdetail to main table.
I have use Corelab MyDac and the speed is correct, but I don't like the core lab ( flowwork) and is very expencive to bay this component for only this issue they don't gave more function than zeoslib.
Eny more thanks for you idias.