Remote shell terminates after 8hours

[BLZB0B]

Hi,

I have an issue where I open a shell:

shell_id = self.protocol.open_shell()

and run a batch file:

command_id = self.protocol.run_command(shell_id, command, args)

One particular batch file I run, regularly takes over 9 hours to complete. however, the shell appears to terminate at exactly 8 hours. (the batch carries on running and completes an hour later).

from 8 hours onwards, when getting the output:

self._raw_get_command_output(shell_id, command_id)

a 400 error is returned. I guessing because the shell has died.

looking at the pywinrm protocol code, work was started to implement the lifetime param:

pywinrm/winrm/protocol.py

Line 154 in b935aca

# if lifetime:

is there a reason this was pulled ? can it be put back easily? would it work?

I could probably do a pull request for this if I could get some background on why it was commented out ?

Phil

[BLZB0B]

after a lot more digging...

https://docs.microsoft.com/en-us/windows/win32/winrm/installation-and-configuration-for-windows-remote-management#maxshellruntime

MaxShellRunTime
Specifies the maximum time, in milliseconds, that the remote command or script is allowed to execute. The default is 28800000.

WinRM 2.0: The MaxShellRunTime setting is set to read-only. Changing the value for MaxShellRunTime will have no effect on the remote shells.

so - no amount of setting the maxshellruntime from the command line will work - it will always default to 8 hours

I suspect if you had an old (winrm 1) install and modified the value to > 8hours, upgrading to winRM 2 or 3 has not reset the stored value to 8hours so still works with >8hour jobs.

the unknown answer is to the question : would enabling the lifetime parameter override this?