It seems that, every so often, dnsmasq needs to be restarted - unless we blame WIDE dhcpv6. Here, dnsmasq has been running for about a week, and a newly started dhcpv6 client will show things like this in the log:
dhcp6c[412]: dhcp6_get_options: get DHCP option identity association, len 18 dhcp6c[412]: IA_NA: ID=0, T1=0, T2=0 dhcp6c[412]: copyin_option: get DHCP option status code, len 2 dhcp6c[412]: status code: no addresses dhcp6c[412]: get_ia: make an IA: NA-0 dhcp6c[412]: update_ia: status code for NA-0: no addresses dhcp6c[412]: update_ia: IA NA-0 is invalidated dhcp6c[412]: remove_ia: remove an IA: NA-0 Restarting dnsmasq always makes things work again, for awhile, and the client log will show instead like: dhcp6c[412]: dhcp6_get_options: get DHCP option identity association, len 40 dhcp6c[412]: IA_NA: ID=0, T1=1800, T2=3150 dhcp6c[412]: copyin_option: get DHCP option IA address, len 24 dhcp6c[412]: copyin_option: IA_NA address: 2604:dead:beef:cafe::01 pltime=3600 vltime=3600 dhcp6c[412]: dhcp6_get_options: get DHCP option status code, len 9 lapis dhcp6c[412]: status code: success I've mentioned this on the mailing list from time to time, but no response yet. If anyone has ideas about how to debug dnsmasq when it begins sending an empty Nontemporary Address, please let me know. Arch Linux dnsmasq 2.76-4 _______________________________________________ Dnsmasq-discuss mailing list Dnsmasq-discuss@lists.thekelleys.org.uk http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss