Level Extreme platform
Subscription
Corporate profile
Products & Services
Support
Legal
Français
Performance issues when querying shared data
Message
General information
Forum:
Visual FoxPro
Category:
Client/server
Environment versions
Visual FoxPro:
VFP 6
Miscellaneous
Thread ID:
01057091
Message ID:
01057892
Views:
11
Thanks Ranjan,

No views in use.

Regards

Mathias

>My first implementation of a VFP application showed severe performance degradation when more users started using the table. After some soul searching the problem was discovered.
>I used updatable views also for reporting. After I created a a prg that wrote the results into a cursor the problems were gone.
>Are you using views? If so, are the updatable?
>
>Regards,
>Ron
>
>>Does the table have an index on DELETED() and, if so, what are the proportion of deleted records versus good records?
>>
>>good luck
>>
>>
>>>Hi Mike,
>>>
>>>Yes I can and I have by introducing a dummy where clause. It optimises and halves the time roughly 2 seconds.
>>>
>>>It would appear optimisation is the real issue. I can live with the slight difference, and I am hoping the customer will be happy with that.
>>>
>>>Cheers!
>>>
>>>Mathias
>>>>Hi Mathias
>>>>
>>>>>Hi All,
>>>>>
>>>>>I have had reports and a colleague has confirmed that a query takes significantly longer if the table being queried is open shared so:
>>>>>
>>>>>Select a.* from Customer a Into Cursor curCustomer where ??? && takes 2 seconds when I am the only user in the system
>>>>>
>>>>>The same query takes 5 seconds if the table customer is open by another user.
>>>>>
>>>>>No views involved
>>>>>
>>>>>Any ideas please?
>>>>
>>>>I just tried a small experiment. Querying a million record table for 10,000 results over the network - only me using the table .591 seconds. With a second user, .7 seconds. Of course my query is fully optimized. Caching will play a role.
>>>>
>>>>I would point out, that testing single user access versus multi-user access is not the issue. Most of the time, there will be multi-user access. The issue then is, can you get the query time down from 5 seconds at all?
Mathias Banda

Time is longer than a rope.
Previous
Reply
Map
View

Click here to load this message in the networking platform