SIGQUIT vs SIGTERM
wouter at nlnetlabs.nl
Wed Oct 14 12:24:54 UTC 2015
-----BEGIN PGP SIGNED MESSAGE-----
On 10/14/2015 01:44 PM, Dag-Erling Smørgrav via Unbound-users wrote:
> Traditionally, Unix daemons will reload their configuration upon
> receiving SIGHUP and terminate gracefully upon receiving SIGTERM.
> Unbound follows this tradition, but in addition, it treats SIGINT
> (Ctrl-C when not daemonized) and SIGQUIT (Ctrl-\ when not
> daemonized) as equivalent to SIGTERM. The man page documents
> SIGQUIT as the correct way to terminate Unbound.
The initial man page was completely wrong, it seems. I did not know
the difference between SIGQUIT and SIGTERM when I wrote it.
> Setting aside the question of wheter intercepting SIGQUIT is a good
> idea (I don't think it is, but it's probably too late to change
> it), this state of affairs seems to confuse users. Consider this
> FreeBSD bug report from a user who claims that Unbound terminates
> cleanly when it receives SIGQUIT, but not when it receives
> In addition, some platforms don't even have SIGQUIT (cf r1097).
> I therefore propose to change the man page so that it mentions
> SIGTERM instead of SIGQUIT:
Of course, I'll change the man-page with the patch.
Best regards, Wouter
> Index: doc/unbound.conf.5.in
- --- doc/unbound.conf.5.in (revision 3504)
> +++ doc/unbound.conf.5.in (working copy) @@ -481,7 +481,7 @@ .fi
> triggers a reload, .nf -kill \-QUIT `cat @UNBOUND_PIDFILE@` +kill
> \-TERM `cat @UNBOUND_PIDFILE@` .fi gracefully terminates. .TP
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1
-----END PGP SIGNATURE-----
More information about the Unbound-users