[Bug 1752411] [NEW] bind9-host, avahi-daemon-check-dns.sh hang forever causes network connections to get stuck

[Launchpad Bug Tracker]

You have been subscribed to a public bug by Trent Lloyd (lathiat):

[Impact]

 * Network connections for some users fail (in some cases a direct
interface, in others when connecting a VPN) because the 'host' command
to check for .local in DNS called by /usr/lib/avahi/avahi-daemon-check-
dns.sh never times out like it should - leaving the script hanging
indefinitely blocking interface up and start-up. This appears to be a
bug in host caused in some circumstances however we implement a
workaround to call it under 'timeout' as the issue with 'host' has not
easily been identified, and in any case acts as a fall-back.

[Test Case]

 * Multiple people have been unable to create a reproducer on a generic
machine (e.g. it does not occur in a VM), I have a specific machine I
can reproduce it on (a Skull Canyon NUC with Intel I219-LM) by simply
"ifdown br0; ifup br0" and there are clearly 10s of other users affected
in varying circumstances that all involve the same symptoms but no clear
test case exists. Best I can suggest is that I test the patch on my
system to ensure it works as expected, and the change is only 1 line
which is fairly easily auditible and understandable.

[Regression Potential]

 * The change is a single line change to the shell script to call host with "timeout". When tested on working and non-working system this appears to function as expected. I believe the regression potential for this is subsequently low.
 * In attempt to anticipate possible issues, I checked that the timeout command is in the same path (/usr/bin) as the host command that is already called without a path, and the coreutils package (which contains timeout) is an Essential package. I also checked that timeout is not a built-in in bash, for those that have changed /bin/sh to bash (just in case).

[Other Info]
 
 * N/A

[Original Bug Description]

On 18.04 Openconnect connects successfully to any of multiple VPN
concentrators but network traffic does not flow across the VPN tunnel
connection. When testing on 16.04 this works flawlessly. This also
worked on this system when it was on 17.10.

I have tried reducing the mtu of the tun0 network device but this has
not resulted in me being able to successfully ping the IP address.

Example showing ping attempt to the IP of DNS server:

~$ cat /etc/resolv.conf
# Dynamic resolv.conf(5) file for glibc resolver(3) generated by resolvconf(8)
#     DO NOT EDIT THIS FILE BY HAND -- YOUR CHANGES WILL BE OVERWRITTEN
# 127.0.0.53 is the systemd-resolved stub resolver.
# run "systemd-resolve --status" to see details about the actual nameservers.

nameserver 172.29.88.11
nameserver 127.0.0.53

liam at liam-lat:~$ netstat -nr
Kernel IP routing table
Destination     Gateway         Genmask         Flags   MSS Window  irtt Iface
0.0.0.0         192.168.1.1     0.0.0.0         UG        0 0          0 wlp2s0
105.27.198.106  192.168.1.1     255.255.255.255 UGH       0 0          0 wlp2s0
169.254.0.0     0.0.0.0         255.255.0.0     U         0 0          0 docker0
172.17.0.0      0.0.0.0         255.255.0.0     U         0 0          0 docker0
172.29.0.0      0.0.0.0         255.255.0.0     U         0 0          0 tun0
172.29.88.11    0.0.0.0         255.255.255.255 UH        0 0          0 tun0
192.168.1.0     0.0.0.0         255.255.255.0   U         0 0          0 wlp2s0
liam at liam-lat:~$ ping 172.29.88.11
PING 172.29.88.11 (172.29.88.11) 56(84) bytes of data.
^C
--- 172.29.88.11 ping statistics ---
4 packets transmitted, 0 received, 100% packet loss, time 3054ms

ProblemType: Bug
DistroRelease: Ubuntu 18.04
Package: openconnect 7.08-3
ProcVersionSignature: Ubuntu 4.15.0-10.11-generic 4.15.3
Uname: Linux 4.15.0-10-generic x86_64
ApportVersion: 2.20.8-0ubuntu10
Architecture: amd64
CurrentDesktop: ubuntu:GNOME
Date: Wed Feb 28 22:11:33 2018
InstallationDate: Installed on 2017-06-15 (258 days ago)
InstallationMedia: Ubuntu 16.04.1 LTS "Xenial Xerus" - Release amd64 (20160719)
SourcePackage: openconnect
UpgradeStatus: Upgraded to bionic on 2018-02-22 (6 days ago)

** Affects: avahi (Ubuntu)
     Importance: High
     Assignee: Trent Lloyd (lathiat)
         Status: Confirmed

** Affects: bind9 (Ubuntu)
     Importance: High
         Status: Confirmed

** Affects: openconnect (Ubuntu)
     Importance: Undecided
         Status: Invalid

** Affects: strongswan (Ubuntu)
     Importance: Undecided
         Status: Invalid

** Affects: avahi (Debian)
     Importance: Unknown
         Status: New


** Tags: amd64 apport-bug bionic server-next
-- 
bind9-host, avahi-daemon-check-dns.sh hang forever causes network connections to get stuck
https://bugs.launchpad.net/bugs/1752411
You received this bug notification because you are a member of Ubuntu Sponsors Team, which is subscribed to the bug report.