-----BEGIN PGP SIGNED MESSAGE-----
On Tue, 20 Jun 2000, Rick Moen wrote:
> Well, but that begs the question of why you would start sshd late in the
> startup process, i.e., in rc.local. On a SysVInit system, one ordinarily
> has a separate startup script, early in the normal runlevel.
I could move sshd up to earlier in the boot process, but I don't think it's
the correct solution for this problem.
As someone else mentioned, sshd could also died later due to any number of
reasons. Even if you have gigabytes of ram, it could happen that a proc
run wild and the system starts killing off procs, and sshd happen to be
one. It'd be nice to have a backup way to get in to fix after the fact.
Running sshd on another port with inetd is a good backup. It's not used
for everyday login, but for those emergencies when it does not matter
if the method is inefficient or slow. As long as I don't have to get
dressed at 2am in the morning and drive a long way in to fix it :-).
> This has the advantage, come to think of it, of avoiding having sshd shut
> down and starting up in lockstep with inetd. (Yes, the currently
> running instance doesn't stop just because inetd does, but that's beside
> the point.)
You've lost me here.
Tin Le
- ----
http://tin.le.org
Internet Security and Firewall Consulting
Tin Le - [EMAIL PROTECTED]
-----BEGIN PGP SIGNATURE-----
Version: 2.6.2i
iQCVAgUBOVE9jhiIIbPkDHhBAQHUWwP/dHkINkLoo1M2Qlpbn0WS05qm2YGNrxJv
HTtnZ7FxIlmOUblWpl01uTIb+cXVhyDJeeXJdBnJ7eIdtpYpk9+9mt6s+OyCNzcj
GxWTgtbFIga1+ujuvZiV/aYbFMOo98mHglMpPqjAs52K3LBR3+xRWMH1xAFKvay2
Ys+IBuT/n3g=
=eazM
-----END PGP SIGNATURE-----
