Well, Larry, I'd start looking at "external" factors then. Things like:

Are their workstations configured with reasonable RAM?...Reasonable CPU (say, gretaer than 333mhz)?

Do the users run much other stuff concurrently with the appl.in question (like Notes/Outlook, some emulation software to mainframe services, etc)?

Do the workstations run anti-virus software of any kind? If so, can it be configured to ignore processing of VFP-related file types like .DBF, .CDX, .SCX, .SCT, .DBT, .DCX, .DCT, etc?

Is their network reasonably configured...low collisions, good buffering in NOS machine (lots of RAM and good HD system) and related? Do they have a network person to regularly review/maintain/tune the network and all of its parts?

Do all of the users practise "safe computing" like sign out of appl. before leaving workstation, clean shutting down and stuff like that?

Barring those things, can you describe a little better what the slowing down looks like? Like... specific transaction types only?... when some user is doing some specific thing?...As the day goes on regardless of number of users?...Only when some special number of users are using the system?
Anything that better describes will be helpful.

again, good luck




>Hi Jim -
>Thanks for the reply
>
>
>You wrote:
>I ASSuME that when you say "large data file sets" you mean large .DBF/.CDX tables.
>Reply: Yes
>
>You wrote:
>What is/are the larger and more heavily used tables and what exact index TAGs are defined in each?
>Reply: the tags are exactly as that of the index names
>
>You wrote:
>Do your WHERE/FOR etc. clauses have expressions coded that match the TAG declarations exactly, as required for Rushmore to operate?
>Reply: Yes
>
>You wrote:
>Are the multi-users all using some 'proper' network/file-server OS, as opposed to WinXX peer-to-peer?
>Reply: They are using Windows NT, each ws is one of the Windows OS
>
>
>You wrote:
>Does the application use lots of SQLs or LOCATEs or SCANs. If so, which?
>Reply: Mainly Seek(alltrim(xxx)), as well as Locates (indexes have been switched off when doing locates), some do while, no scans
>
>You wrote:
>Are you deliberately locking records or are you leaving tht to VFP to handle?
>Reply: Letting VFP handle it
>
>You wrote:
>What type of buffering/locking is in use in the application?
>Reply: Optimistic record
>
>I thought as much, when we run it here in our NT environment, it seems to process quite well, however, we do not have as many users as our user. Nor as many records. The largest record set of 888K
>
>Thanks,
>
>Larry