Plateforme Level Extreme
Abonnement
Profil corporatif
Produits & Services
Support
Légal
English
25meg. network package for a SQL-12k file result-set
Message
 
À
16/03/2002 19:36:24
Information générale
Forum:
Visual FoxPro
Catégorie:
Base de données, Tables, Vues, Index et syntaxe SQL
Divers
Thread ID:
00633822
Message ID:
00633839
Vues:
16
>>>>What is wrong here?
>>>>
>>>>Have data (FoxPro tables) on a remote server and when executing a SQL through a .EXE ( using a stored procedure ) the resultant is 50 records equaling about a file size of 12K dbf yet we see over 25 megs. coming over the network? It wasn't until we put a sniffer on the network because of slow performance that this came to view. Comparing this to execting a SQL from the application showed no detecable difference. I was always under the impression that using stored proceedures only the resultant record set would be delivered and for that manner the same when execting a SQL from the client side.
>>>
>>>In the file server setup everithing is executed on the client side including VFP stored procedures.
>>
>>
>>Is there anyting that can be done to have the server only deliver the resultant?
>
>Hi Bob
>
>Your original aaumption would hold if you were using SQL Server. But using Fox tables across the network, there is no mechanism to execute the filtering process on the remote drive, and return only the result.
>
>By default, Your search will be Rushore optimized, and more often than not, it requires the the complete index set(s), which in your case makes up the 25 megs you see comimg across the wire. If you know how the values are distributed across the tables, it may be possible to fool Rushmore into processing only a subset of the indexes, but it may be difficult to be wise before the event.
>
>Where your tables are on a network, and 'large', there are two things you can try, which may dramatically improve performance.
>
>1. Use SET KEY, with or without without filters. I implemented this a few months ago wherever there was a bottleneck on grids.
>
>2. SET OPTIMISE OFF (turn off Rushmore) I experimented with this only yesterday, testing for and turning off the optimise if the drive on which the DBC in which the tables are contained is a network drive.
>
>I found in both cases that there was a vey dramatic improvement on large tables.
>
>Geoff


Thanks Geoff.. I also sent you a reply via Email asking for further help.
Précédent
Répondre
Fil
Voir

Click here to load this message in the networking platform