dns lookup fails [was: doesn't fall back to ipv4]

asked 2016-01-20 10:46:18 +0300

dmnk gravatar image

updated 2016-01-27 07:23:36 +0300


  • routes are set up correct (ping a public IP works)
  • dns-lookups fail for WLAN = enabled
  • connman service has the correct namserver ip set, but it seems to be unused
  • as soon as WLAN is deactivated, dns works again


how can i debug the DNS connection, as nslookup and nmcli are missing?

full story

seems that i might have bin mislead by my setup. now it looks like networkmanager is ignoring the DNS provided by DHCP{4,6} at all. just the specific conditions remain unclear.

My network is (sometimes) linked to the internet via 6to4, but the local connection is always dual-stack. so from time to time IPv6 works over the LAN boundary, sometimes not.

If the network is "uplinked" via v6, the router indicates this with the "default router" flag set in the router advertisements. So there shouldn't be a default route available for sailfish to use otherwise.

till this wasn't a problem, but since the update the dns resolution fails. i guess its stuck at trying via IPv6, which doesn't provide a path to the internet right now.

as soon as i deactivate WLAN and am only on the mobile network connection, the dns client (and therefore mail, browser, store, ...) come to live again.

edit retag flag offensive close delete


Can you get output of "ip -6 route" when the issue is happening?

tigeli ( 2016-01-20 23:28:35 +0300 )edit

` [nemo@Jolla ~]$ /sbin/ip -6 route

fdbf:52fc:819a::1 via fdbf:52fc:819a::1 dev wlan0  metric 0 


fdbf:52fc:819a::/64 dev wlan0  proto kernel  metric 256  expires 7156sec

unreachable fe80::/64 dev lo  proto kernel  metric 256  error -101

fe80::/64 dev wlan0  proto kernel  metric 256`
dmnk ( 2016-01-21 09:24:36 +0300 )edit

as addition: the active connections via v4 and 6 while ping is hanging on another session

` [nemo@Jolla bin]$ ./netstat -n -A inet6 Active Internet connections (w/o servers) Proto Recv-Q Send-Q Local Address Foreign Address State
tcp 0 0 ::ffff: ::ffff: ESTABLISHED tcp 0 320 ::ffff: ::ffff: ESTABLISHED tcp 0 0 ::ffff: ::ffff: ESTABLISHED udp 0 0 fdbf:52fc:819a:0::35075 fdbf:52fc:819a::1:53 ESTABLISHED udp 0 0 fdbf:52fc:819a:0::58635 fdbf:52fc:819a::1:53 ESTABLISHED udp 0 0 fdbf:52fc:819a:0::52584 fdbf:52fc:819a::1:53 ESTABLISHED [nemo@Jolla bin]$ ./netstat -n -A inet Active Internet connections (w/o servers) Proto Recv-Q Send-Q Local Address Foreign Address State
udp 0 0 ESTABLISHED udp 0 0 ESTABLISHED [nemo@Jolla bin]$ /sbin/ifconfig lo Link encap:Local Loopback

wlan0 Link encap:Ethernet HWaddr 50:56:A8:01:69:84
inet addr: Bcast: Mask: inet6 addr: fdbf:52fc:819a:0:5256:a8ff:fe01:6984/64 Scope:Global inet6 addr: fe80::5256:a8ff:fe01:6984/64 Scope:Link UP BROADCAST RUNNING MULTICAST DYNAMIC MTU:1500 Metric:1 RX packets:910 errors:0 dropped:0 overruns:0 frame:0 TX packets:1197 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:100 RX bytes:82569 (80.6 KiB) TX bytes:203200 (198.4 KiB) `

dmnk ( 2016-01-21 09:25:00 +0300 )edit

i had some troubles reproducing the behaviour, so i can add the error condition:

  • if at the time of starting the phone WLAN and mobile internet connection have been enabled.

didn't happen if just WLAN was enabled or if i set both to enabled afterwards.

dmnk ( 2016-01-21 09:29:58 +0300 )edit

I have tried at home and works as expected degrading to IPv4:

[root@Jolla nemo]# curl -Iv google.com
* Rebuilt URL to: google.com/
*   Trying 2a00:1450:4003:807::200e...
*   Trying
* Connected to google.com ( port 80 (#0)
itorres ( 2016-01-21 17:57:14 +0300 )edit