Message ID | 82fd5037dd1ec543b4df97995bb55d61f5165334.1688515814.git.lilah@lunabee.space |
---|---|
State | New |
Headers | show |
Series | [bug#64468] services: dhcp-client-service-type: Support DDNS. | expand |
Hi, Lilah Tascheter <lilah@lunabee.space> skribis: > * gnu/services/networking.scm (dhcp-client-shepherd-service): Enable -I > flag on dhclient. [...] > + (cons* dhclient "-nw" "-I" Could there be unintended side effects from passing “-I”? The dhclient(8) man page is silent about the implications. If there are none, the patch LGTM. Thanks, Ludo’.
Ludovic Courtès <ludo@gnu.org> wrote:
> Could there be unintended side effects from passing "-I"?
the isc-dhcp knowledgebase says that by default dhclient uses a
pre-standardization implementation of DDNS, which is incompatable with non-isc
DHCP servers. no idea why that's the default, but -I just switches to using the
standardized protocol.
kb link with more details: https://kb.isc.org/docs/aa-01091
~lunabee
Lilah Tascheter <lilah@lunabee.space> skribis: > Ludovic Courtès <ludo@gnu.org> wrote: >> Could there be unintended side effects from passing "-I"? > > the isc-dhcp knowledgebase says that by default dhclient uses a > pre-standardization implementation of DDNS, which is incompatable with non-isc > DHCP servers. no idea why that's the default, but -I just switches to using the > standardized protocol. > > kb link with more details: https://kb.isc.org/docs/aa-01091 Thanks for explaining. I added this explanation as a comment in the code and committed. Ludo’.
diff --git a/gnu/services/networking.scm b/gnu/services/networking.scm index 5657b141d9..0fbba1b8c7 100644 --- a/gnu/services/networking.scm +++ b/gnu/services/networking.scm @@ -350,7 +350,7 @@ (define ifaces (false-if-exception (delete-file #$pid-file)) (let ((pid (fork+exec-command - (cons* dhclient "-nw" + (cons* dhclient "-nw" "-I" "-pf" #$pid-file ifaces)))) (and (zero? (cdr (waitpid pid))) (read-pid-file #$pid-file)))))