I am experiencing the same lost of connectivity on my configuration. Large continuous file transfers from one of my Windows Server 2k3 VM's to my Solaris 10 (2009/10) install causes the Solaris networking adapter to become unresponsive. Halting the VM, waiting a minute or so, and restarting it brings it back up. This happens every time I start to do a large file transfer. Normal network traffic (SSH, HTTP) don't trigger this behavior.
Here are some details from my configuration:
SuperMicro X7DB8
Dual quad-core Xeon E5405's, 16GB RAM, 2TB storage on RAID5
Two Server 2k3 VM's, one Server 2k8, one Ubuntu Server 9.04, one Solaris 10 (2009/10)
All VM's are configured using the stock Intel Pro1000 adapters. There are two virtual switches configured for internal and external connectivity. The failure occurs when transferring data from the Ubuntu Samba server through one of the 2k3 servers to the CIFS share presented by the Solaris VM. I haven't verified that this failure occurs when using something like SCP directly from the Ubuntu VM to the Solaris VM.
*** ADDENDUM: File copies over SCP don't appear to be presenting any issues. I haven't tried rsync yet, but my guess is that anything that tunnels through SSH will work ok.{color:#ff0000} ***{color}
