Client in virtual machine looses connection. Max values for Ping or disabling them?

Submitted by plme on #1

Hello everyone,

 I trying to work VirtualHere into our organization which by and large works fine. One problem though is with our virtual machines. They can reach the server fine, direct connect with IP, and check out a dongle but it gets returned quick with a "server timeout" message. This happens both with the virtual NICs in NAT and bridged mode.

 I think this is related to the server wanting to ping the clients. When using bridge mode in a vm the server can ping the vm but VirtualHere still times out.

 Using the "reverse clients" function is as I understand it unworkable for us, we need something that works on the fly without extra administration when machines change.

My questions:

 Have anyone here used VirtualHere successfully in a virtual machine and if so how did you get it to work?

 What are the max values for PingInterval and PingTimeout?

 Can you turn off the server trying to ping clients? If so, how?

Submitted by Michael on Tue, 2024/03/05 - 07:13

#2

VirtualHere works fine and the pings need to stay as they are. I think you have some network issue. Virtualhere uses TCP 7575 and sends a ping in both directions every 3 seconds.

Submitted by plme on Tue, 2024/03/05 - 17:29

#3

I'm confident the network is good.

Why is the server pinging the clients? This makes no sense to me. Does the server try to ping the client IP or hostname?

Raising the server ping values does so far seem to resolve the issue so keeping them where they are as default appears not to workable.

Submitted by Michael on Tue, 2024/03/05 - 18:12

#4

The client and server need to ping each other otherwise there is no way to determine if the connection is dropped.

Submitted by plme on Tue, 2024/03/05 - 22:26

#5

That's the bit that doesn't make sense for me. That the clients regularly send a "I'm client X and I'm still using dongle Y" to the server I understand. Likewise that if the server doesn't receive an "I'm still using..." message for a given time it would consider the connection broken and free up the resources used by said client.

Looking at the FAQs and browsing this forum the requirement for them to have contact with each other both ways seems to create problems that on the face of things could be avoided. I know that this is second guessing years of development but for a client-server setup it makes no sense to me, peer-peer would be different.