My company has an extensive, mission critical VFP 9 application in use in numerous sites. One client has recently upgraded to Server 2008 R2, and running it numerous VM's (were running XP, now running Server 2008) and numerous TS sessions. We are not explicitly limiting RAM usage (Sys(3050)) yet, but think the VMs may have a limit. We use a standard VFP database, with extensive use of BEGIN and END transactions for data changes, no buffering.
There appears dramatic slowing of the application at random locations, typically around record replaces. ContrAcct had the same question a while back but a non-specific solution was forced. We use minmal specific record locking but acknowledge this may be a factor in this environment.
Virus control is turned off, machines are all local at the site.
Has anyone had real problems around this and identified real specific solutions?