We have a situation in one of our customer where a 3rd party medical application hangs at times and the services on the application server need to be restarted to continue using the application. It seems that all the clients connection has been droped from the application server. This apps runs in 2 tier environment with 1 database and 1 application server which are on vSphere. The app vendor that wrote the app somehow is putting the blame on the VM environment because this issue did not happen when runnning on physical machine previously. All the installation are made from scratch on the VMs.
The databse is running on SQL 2005 Standard with 2 cpu and 8G of ram and the application server is running from PowerBuilder v5 on 2 cpu with 4G of ram VM. The VMs are running on Win2k3 enterprise 32bit OS and connected through e1000 vNIC.
Initialy this VMs runs on a shared Vswitch with the other VMs. After the complaint we have allocate this VMs to a dedicated nic port on a different vswitch prepared for this particular VM only, the occurance has decrease. Based on the Virtual Center performance chart this VMs are running max 60 and average 25 to 30 percent utilization. The ESXs already patched with latest critical patches.
Is there any way to show that this issue are not caused by the MVs or it's network performance or resources.
