[FreeNX-kNX] Bug: negotiating link parameters -> timeout ON resume session. v.0.6.0/2.1.0-17

[torrr com]

Hi,

  I am not sure if the trees would keep their format through the
e-mail procedure, so I have shortened them by calling one common tree
DTree, and an other common tree ETree. BTW both of these trees are
very similar with the difference that the cat(29276) and
nxserver(29261) exist in DTree, and are trimmed at ETree.

DTree, and ETree definitions:
---------------------------------------------------------------------------------
DTree=
nxserver(29256)-+-nxserver(29257)-+-cat(29276)
                |
`-nxserver(29273)---nxnode-login(29275)-+-ssh(29278)
                |
   `-{nxnode-login}(29277)
                `-nxserver(29261)

---------------------------------------------------------------------------------
ETree=
nxserver(29256)---nxserver(29257)---nxserver(29273)---nxnode-login(29275)-+-ssh(29278)

   `-{nxnode-login}(29277)
---------------------------------------------------------------------------------

The test procedure:
::::::::::::::
FirstLogin
::::::::::::::
sshd(29105)---netcat(29106)---DTree

::::::::::::::
FirstSuspend
::::::::::::::
DTree
sshd(29105)          #comment: Its the first SSH process.

::::::::::::::
FirstSuspendResumeB4Crush
::::::::::::::
DTree
sshd(29105)            #comment: Its the first SSH process.
sshd(29689)---netcat(29690)---nxserver(29910)

::::::::::::::
FirstSuspendResumeAfterCrush
::::::::::::::
DTree
sshd(29105)
sshd(29689)             #comment:
                                   # This looks just like the
privious, just without the ---netcat(...

::::::::::::::
FirstSuspendResumeAfterCrushNXSSH.exeTerminated
::::::::::::::
DTree                       #Both sshd disappeared.

::::::::::::::
FirstSuspendResumeAfterCrushNXSSH.exeTerminatedResume
::::::::::::::
ETree
sshd(30402)---netcat(30403)---nxserver(30729)

::::::::::::::
FirstSuspendResumeAfterCrushNXSSH.exeTerminatedResumeSuspend
::::::::::::::
ETree

::::::::::::::
FirstSuspendResumeAfterCrushNXSSH.exeTerminatedResumeSuspendResume
::::::::::::::
ETree
sshd(31036)---netcat(31037)---nxserver(31257)

::::::::::::::
FirstSuspendResumeAfterCrushNXSSH.exeTerminatedResumeSuspendResumeSuspend
::::::::::::::
ETree

::::::::::::::
FirstSuspendResumeAfterCrushNXSSH.exeTerminatedResumeSuspendResumeSuspendResume
::::::::::::::
ETree
sshd(31586)---netcat(31587)---nxserver(31807)


-----------------------------------------

I will try again, and this time kill the first sshd process from the
server, after the first suspend is done, and see if it helps. Here
this first sshd was sshd(29105). Yep, it helped , no problem with
resume. Checking the trees, suspending checking the trees, they look
like the good trees at the end of the test. OK, so the solution might
be killing the first sshd at the server right after the first suspend.

PS. If I can edit something somewhere and fix it, I'd love to get
instructions/pointers.