Reinstate keepalive control

[bitprophet]

v1 had a simple integer env.keepalive / --keepalive that got turned into client.get_transport().set_keepalive(xxx); v2 has not ported that over yet.

[dgskwirut]

Any update on if this can be ported soon? Seems like a simple fix.

[rdenham]

I'd also like this to be implemented.

[grokpot]

How would one implement this functionality currently with Fabric2? Or, is it even possible?

[cr-liadsh]

This might be a dirty workaround, but it might be possible to set the timeout this way:
from fabric2 import Connection
c = Connection(host="1.2.3.4", user="user", connect_kwargs={"key_filename": "my_key"}
c.run("echo hey")
c.transport.set_keepalive(<keepalive>)

I didn't find a different way to set keepalive with fabric2 at the moment ):

[ibrahimasow1]

@cr-liadsh thank you for your workaround!
I was stuck for weeks and couldn't figure out why my commands were hanging. Setting the keep alive allowed me to get the command's result.
The port of Fabric v1's keepalive to Fabric v2 is still pending, so I will keep using your fix.