On Wed, Feb 01, 2017 at 05:09:43PM +0200, Lars Noodén wrote: > On 02/01/2017 05:06 PM, Erling Westenvik wrote: > > On Wed, Feb 01, 2017 at 03:58:51PM +0100, Manuel Giraud wrote: > >> Erling Westenvik <erling.westen...@gmail.com> writes: > >> > >>> However, I got inspired and when I disabled pf (pfctl -d) I got full > >>> contact! (But -- when I turned pf back on (pfctl -e) I lost the one > >>> connection I had... Now I have to wait 48 minutes for the server to > >>> reboot. Not much more to do now except for crossing my fingers...) > >> > >> Err, yes but won't pf be enabled at boot time? Hopefully, some of your > >> pf tables will be reset. > > > > True. But before I turned pf off and back on, I couldn't be sure what > > was causing the problem. If it was an external problem I would've been > > better off with the one active existing ssh connection. > > > > I hope it reboots ok. If you end up with a similar situation again you > might set up 2 or more at jobs that build reverse tunnels from port 22 > to an outside machine. That way you can still connect back via a tunnel > if the main SSH session drops. That won't solve the problem but might > buy you more time to investigate. > > /Lars
Thanks. Good advice and I actually HAD such a tunnel to a remote machine earlier to day but took it down due to experimenting, but without remembering to reactivate it. Won't forget that again! -- Erling Westenvik
