[ale] After 15 years, Nohup is sttll broken???

[Chris Fowler]

> From: neal at mnopltd.com
> To: "Chris Fowler" <cfowler at outpostsentinel.com>
> Cc: "Atlanta Linux Enthusiasts" <ale at ale.org>
> Sent: Wednesday, July 31, 2019 10:45:20 AM
> Subject: Re: [ale] After 15 years, Nohup is sttll broken???

> Thank you for the reply. Additional response inline below.

> On 2019-07-31 09:19, Chris Fowler wrote:
> > -------------------------

> >> FROM: "Neal Rhodes via Ale" <ale at ale.org>
> >> TO: "Jim Kinney" <jim.kinney at gmail.com>
> >> CC: "Atlanta Linux Enthusiasts" <ale at ale.org>
> >> SENT: Wednesday, July 31, 2019 10:10:41 AM
> >> SUBJECT: Re: [ale] After 15 years, Nohup is sttll broken???

> >> Well, indirectly - if the ssh session didn't die, it probably
> >> wouldn't
> >> happen.
> > I've been using keepalive SSH options for years to keep session up.
> > I had system that did dialout ppp to servers with an idle timeout set
> > for pppd. SSH session would drop so I used the keep alive out of band
> > SSH messages to keep those connections up. I now use the same options
> > for standard networks (real speeds).

> Hmm. I forgot about that. However, the man page for ssh_config states
> that tcpkeepalive is on by default. So...

> I think we're seeing the ssh sessions die because the Sonicwall VPN is
> configured to kill any connections after xx time. I cannot get that
> changed.

Dumb NAT firewall. SSH will NOT send any traffic unless there is traffic to send. If SSH is your only connection then the dumb NAT firewawll will only see packets duing standard TCP keep alive or when SSH sends data. To fix that firewall you need to tell SSH to send data! 

ssh -o ServerAliveInterval=20 -o ServerAliveCount=5 user at host 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mail.ale.org/pipermail/ale/attachments/20190731/c26cf05f/attachment.html>