>I have a client with a large network 25+ workstations, running a vertical market application.
>
>They are having speed issues with their network - not just my application, but general speed.
>
>Their network specialist have advised;
>
>"We have installed WireShark on a workstation and on the server to try and find the cause of the problem. It was noted that there are a lot of ‘Fragmented Packets’ going between the workstation and the server.
>
>We isolated the source of the ‘Fragmented Packets’ by excluding all networked devices and creating a restricted network consisting of only the broadband router, the server and the workstation connected via a Gigabyte switch. Initially there was no ‘Fragmented Packets’, however, as soon as "VFP application" was run the ‘Fragmented Packets’ started again.
>
>It seems clear to us that there is a problem with the way "VFP application" is communicating with the network and therefore we need to work together to try and resolve the problem."
>
>I am not a network specialist, does anyone know how VFP woudl cause fragmented packages and why this would be an issue?
>
>Is there anything I could do to resolve/change this?

You've already received lots of good advice to follow up on.

There are some rare circumstances where a VFP application might cause unusual network behaviour:

- if you're using API calls or 3rd party ActiveXs or FLLs to directly manipulate network packets
- if some other component you're using is doing so, without your knowledge e.g. (potentially) network dongles, copy protection, file/table encryption etc.

Other than that, VFP operates at the Server Message Block (SMB) level and has no influence on the mechanics of how individual packets are transferred across the network.

One thing that may not have been yet said is that the "network specialist" is almost certainly confusing cause and effect. VFP using native tables/databases is a file-server database and generates a lot of network traffic. This traffic will cause network configuration issues to appear, that might not manifest at low traffic levels. You can think of a networked, multi-user VFP app as an excellent "stress test" for an SMB network.

VFP is not causing the problem, it's creating conditions where some other problem or misconfiguration is manifesting.

Another thing to bear in mind is that, if poor performance is the problem, then there is at least one bottleneck somewhere. The "network specialist" has found what may be an unusual condition, but hasn't shown that it's a bottleneck. If it isn't a bottleneck, the problem is elsewhere and you'll need to find it. Try to keep a high-level view of the actual problem ("slow performance") rather than obsessing over symptoms. There are many potential causes of slow network performance.