BUG: Can't reconnect after connection timed out

[eGaTS]

Viewer: v1.0.8.2 on Win7(x64) w/ MSRC4
Server: v1.0.8.2 on XP pro w/ MSRC4
^-- Please note this issue is not OS specific. I just posted that so no one would ask.


This bug is all over these forums, but as far as I can tell it has never been resolved, or the solutions offered are unacceptable.

It goes something like this:
1. User successfully connects to server.
2. User gets distracted, forgets about ultravnc.
3. The connection times out, closing the viewer.
4. The viewer refuses to reconnect because the server still thinks the session is active, and only a single session is allowed.

Status stuck at: "Negotiate protocol version..."
Error message is: "Connection failed - Error reading Protocol Version"


UNACCEPTABLE "solutions" include:
1. Switching to a different encryption plugin.
Why not?: I like MSRC4. The issue is with the server, not the plugin. The reason it occurs with this plugin is that only one session is permitted. That is by design.

2. Don't get distracted.
Why not?: Hey, it happens. You fell asleep, got a phone call, etc.

3. Use some kind of keepalive
Why not?: If the terminal is unattended, it should close the session for security reasons.

4. Implement some kind of wacky scripts or some crap that restarts the server to clear the stuck session, or even the entire machine. This includes using remote programs to restart things.
Why not?: I have my firewall set up to only allow access from ultravnc and I want to keep it that way. As for the scripts, I don't want extra processes running. This should be handled by the server itself.

^-- posting these "solutions", or links to threads containing these "solutions" will be ignored.


The solution we're looking for is one where the server recognizes the connection has been lost, and closes the session, allowing a new one to be created.

If someone has found a server setting that fixes this, please share it (and sticky this thread-- or add it to the FAQ --it truly is asked frequently). If there is no current solution, I ask that the developer please prioritize this bug. I went away on vacation for two weeks, and this bug showed its ugly head on the second day. I was stuck without any access to my machine or data for the duration of the trip. Not only that, but I'm guessing XP didn't lock the server machine either (will check when I go in on Monday). This means anyone in the building had all week to mess around with my machine


Edit: If no solution is found, this should be moved to the feature request board.

[B]

You're coming across as rather coarse and demanding, but in a way I appreciate your summary of this issue (if it's accurate). It might explain some things.

Personally I think the server should be performing its own keepalive checks. The whole session is a two-way handshake, after all.

This may be one of lots of reasons it pays to have more than one way to access your target servers. I had similar issues tunneling SSH at one point.

[eGaTS]

My apologies. I was already in a crappy mood and then got into an argument with someone about this. While having a second remote access method would have given me what I needed this week, it still doesn't fix the bug that has been around for at least four years (based on other forum posts). That is a long time for something this critical. I compare it to the brakes failing on my vehicle, and someone suggesting that I install airbags or a better seat belt to fix the issue. Know what I'm saying?