I want my VMs to be assigned static addresses by hostname. I would prefer to set up a generic way to configure dnsmasq, since I don't think this is the last thing I'll want to change about address assignment. I saw this patch from "Mads Chr. Olesen" https://www.redhat.com/archives/libvir-list/2008-April/msg00327.html But it seems like a lot of work to give access to only one dnsmasq option. I'm working on a patch that enables a dnsmasq.d directory where you can (optionally) put individual configuration files. I have not fully tested it yet, so please don't consider this an official submission. I also don't feel that I have spent enough time with the code to understand the 'philosphy' if you will, the guiding attitudes. Is the all-XML approach better according the libvirt developer community standards? If so, why? Thanks. Untested patch follows. Please let me know if I'm on the right path and if not, why. --Kurt diff -u /root/libvirt-0.4.4-virgin/src/qemu_driver.c ./qemu_driver.c --- /root/libvirt-0.4.4-virgin/src/qemu_driver.c 2008-06-12 07:52:53.000000000 -0700 +++ ./qemu_driver.c 2008-07-13 11:31:26.000000000 -0700 @@ -1048,6 +1048,7 @@ 1 + /* --bind-interfaces */ 2 + /* --pid-file "" */ 2 + /* --conf-file "" */ + 2 + /* --conf-dir DNSMASQCONF_DIR */ /*2 + *//* --interface virbr0 */ 2 + /* --except-interface lo */ 2 + /* --listen-address 10.0.0.1 */ @@ -1081,6 +1082,9 @@ APPEND_ARG(*argv, i++, "--conf-file"); APPEND_ARG(*argv, i++, ""); + APPEND_ARG(*argv, i++, "--conf-dir"); + APPEND_ARG(*argv, i++, DNSMASQCONF_DIR); + /* * XXX does not actually work, due to some kind of * race condition setting up ipv6 addresses on the diff -u /root/libvirt-0.4.4-virgin/src/qemu_driver.h ./qemu_driver.h --- /root/libvirt-0.4.4-virgin/src/qemu_driver.h 2008-04-23 07:19:02.000000000 -0700 +++ ./qemu_driver.h 2008-07-13 11:26:47.000000000 -0700 @@ -33,6 +33,8 @@ int qemudRegister(void); +#define DNSMASQCONF_DIR SYSCONF_DIR "/dnsmasq.d" + #endif /* WITH_QEMU */ #endif /* QEMUD_DRIVER_H */ -- ---------------------------------------------------------------------- Merlot Research Group, Inc kls[at]merlot.com "Quality is free, but only to those who are willing to pay heavily for it." (Lister, DeMarco: "Peopleware")