DNS resolution dropouts

Hello!

Did anything change in relation to DNS and/or connman in the latest update by any chance?

The reason I ask is that since the last update i’ve been experiencing intermittent DNS resolution failures. Everything runs fine for a while (but no more than a day or so), but then DNS falls over and many good things stop working as a result. I managed to catch it in the act just now, and it looks like resolv.conf is being overwritten with a file that contains no DNS servers for some reason. Restarting the connman service (or a reboot, obviously) fixes the problem.

Has anybody else experienced the same thing, or can anybody suggest a fix?

I’m running an NFS root and DHCP IP configuration on a Pi 2, with a remote MySQL database. I realise NFS and DHCP isn’t necessarily a recommended combination, but my network has been stable enough that it hasn’t caused a problem for several years now. My DHCP scope lease is set to 2 days, and the problem is occurring more often than that.

/boot/cmdline.txt

root=/dev/nfs nfsroot=192.168.0.6:/export/pi1,vers=3 ip=dhcp 0.0.0.0:255.255.255.0:osmc:eth0:off rootwait quiet osmcdev=rbp2

Logs (from the first error):

19:37:01 86420.234375 T:1259594736 ERROR: Unable to open database: MyMusic56 [2005](Unknown MySQL server host ‘db.obfuscated.au’ (2))
19:37:01 86420.273438 T:1259594736 ERROR: Previous line repeats 6 times.
19:37:01 86420.273438 T:1259594736 ERROR: SQL: [MyMusic56] The MySQL server has gone away
Query: SELECT songview., songartistview. FROM (SELECT idAlbum FROM album ORDER BY idAlbum DESC LIMIT 10) AS recentalbums JOIN songview ON songview.idAlbum = recentalbums.idAlbum JOIN songartistview ON songview.idSong = songartistview.idSong ORDER BY songview.idAlbum desc, songview.itrack, songartistview.iOrder
19:37:01 86420.273438 T:1259594736 ERROR: GetRecentlyAddedAlbumSongs failed
19:37:01 86420.281250 T:1259594736 ERROR: Unable to open database: MyMusic56 [2005](Unknown MySQL server host ‘db.obfuscated.au’ (2))
19:37:01 86420.289062 T:1259594736 ERROR: Previous line repeats 4 times.
19:37:01 86420.289062 T:1259594736 ERROR: SQL: [MyMusic56] The MySQL server has gone away
Query: SELECT albumview., albumartistview. FROM (SELECT idAlbum FROM album WHERE strAlbum != ‘’ ORDER BY idAlbum DESC LIMIT 10) AS recentalbums JOIN albumview ON albumview.idAlbum = recentalbums.idAlbum LEFT JOIN albumartistview ON albumview.idAlbum = albumartistview.idAlbum ORDER BY albumview.idAlbum desc, albumartistview.iOrder
19:37:01 86420.289062 T:1259594736 ERROR: GetRecentlyAddedAlbums failed
19:37:01 86420.312500 T:1259594736 ERROR: Unable to open database: MyVideos99 [2005](Unknown MySQL server host ‘db.obfuscated.au’ (2))
19:37:01 86420.351562 T:1259594736 ERROR: Previous line repeats 6 times.
19:37:01 86420.351562 T:1259594736 ERROR: SQL: [MyVideos99] The MySQL server has gone away
Query: select * from movie_view ORDER BY dateAdded desc, idMovie desc LIMIT 10
19:37:01 86420.351562 T:1259594736 ERROR: GetMoviesByWhere failed
19:37:01 86420.359375 T:1259594736 ERROR: Unable to open database: MyVideos99 [2005](Unknown MySQL server host ‘db.obfuscated.au’ (2))
19:37:01 86420.367188 T:1259594736 ERROR: Previous line repeats 4 times.

Troubleshooting:

osmc@Kodi-L:~$ ping db.obfuscated.au
ping: bad address ‘db.obfuscated.au’
osmc@Kodi-L:~$ ping www.google.com
ping: bad address ‘www.google.com
osmc@Kodi-L:~$ ls -al /etc/resolv.conf
-r–r–r-- 1 root root 34 Jan 8 19:36 /etc/resolv.conf
osmc@Kodi-L:~$ cat /etc/resolv.conf
# Generated by Connection Manager
osmc@Kodi-L:~$ sudo systemctl restart connman
osmc@Kodi-L:~$ cat /etc/resolv.conf
# Generated by Connection Manager
nameserver 192.168.0.10
nameserver 192.168.0.254
nameserver 8.8.8.8
osmc@Kodi-L:~$

As you can see from the timestamps, resolv.conf was updated immediately prior to DNS resolution failing. Any suggestions or advice would be most welcome. Its entirely possible the problem is my own doing somehow, but it did seem to start happening around the time of the last update, so I figured it was worth asking the question.

Many thanks!

Maybe related?

1 Like

Yep. That looks like the same problem. Thanks very much for such a quick response. And apologies for not finding that thread myself. I did search before posting, but clearly not very well.

I’ll implement the /proc/net/pnp workaround and sit tight for the January update.

Many thanks for the help! :slight_smile:

I encounter exactly the same problem, all my support kodi put me the same message: broken pipe
Be it my tablet, my box android
What is surprising is that my pc has access to osmc without any problems.

Sorry for my english, google translate to be understand :slight_smile:

@thanos76 if you are not using NFS root then the issue is not related

Yet it follows the update today

1 Like

There was an update a couple of weeks ago, but not today.
If you are not using an NFS root, then the issue detailed above doesn’t apply to you.

Unfortunately as you have posted no logs we can’t provide any answers.

For some time I do not use osmc as a nas. That’s why I only saw the update today:) that’s how I noticed the problem by not succeeding To have access to the rpi hard disk.
I’ll try to figure out how to get the paper and tell you that again.
Thanks again for your help and your work.