Main Contents

VMware and RDP conflicts

March 6, 2008

I was having the strangest problem recently while trying to RDP into a few different servers. I would get the Windows logon prompt, but after logging in the screen would go black and eventually drop the connection entirely. This looked like MTU problems, which I’ve seen manifested in this way, but this wasn’t the case. Some servers were on a PPPoA DSL connection which allows for a MTU size of 1500, other servers were on a PPPoE connection, but I verified either the MTU size was 1492 or the clamp-mss-to-pmtu option was enabled on the firewall. To add to the strangeness, it only happened from some of my computers, but not my kids PC, for example. Even stranger, I could RDP to some remote servers at one site, but not another at the same site. Packet captures showed some minor problems, such as some DUP ACK’s, but nothing major. The capture ends up with a completely unresponsive remote host. Also, all traffic, even pings, to other devices would stop for about 10 seconds. It turns out that it has to do with connecting to console session (mstsc.exe /console) with RDP to a VMware host server. This apparently can cause all network activity on the guests to drop for a short time. These particular sites I was having a problem with were running their firewall in a virtual machine, so the process of logging into session 0 via RDP killed the firewall connection, dropping all in- and outbound traffic to the entire site for about 10 seconds. Too bad, especially since the VMware console software prefers to run in session 0. Looks like I’ll have to start using VNC for these situations.

Filed under: Microsoft, Network, Windows |

Sorry, the comment form is closed at this time.