** Changed in: bind9 (Ubuntu Cosmic)
Status: Confirmed => Won't Fix
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1752411
Title:
bind9-host, avahi-daemon-check-dns.sh hang forever causes
** Changed in: avahi (Debian)
Status: New => Fix Released
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1752411
Title:
bind9-host, avahi-daemon-check-dns.sh hang forever causes network
In https://bugs.launchpad.net/ubuntu/+source/bind9/+bug/1797926 looks
like the host command, that doesn't timeout and is the subject of this
bug here, crashed.
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
In https://bugs.launchpad.net/ubuntu/+source/bind9/+bug/1797926 looks
like the host command, that doesn't timeout and is the subject of this
bug here, crashed.
--
You received this bug notification because you are a member of Ubuntu
Server, which is subscribed to the bug report.
** Tags removed: server-next
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1752411
Title:
bind9-host, avahi-daemon-check-dns.sh hang forever causes network
connections to get stuck
To manage
** Tags removed: server-next
--
You received this bug notification because you are a member of Ubuntu
Server, which is subscribed to the bug report.
https://bugs.launchpad.net/bugs/1752411
Title:
bind9-host, avahi-daemon-check-dns.sh hang forever causes network
connections to get stuck
To
This bug was fixed in the package avahi - 0.7-3.1ubuntu1.1
---
avahi (0.7-3.1ubuntu1.1) bionic; urgency=medium
[ Trent Lloyd ]
* debian/avahi-daemon-check-dns.sh: On some hardware, the 'host'
command gets stuck and does not timeout as it should leaving this script
and
Package: avahi-daemon
Version: 0.7-3.1ubuntu1.1
Works. Thank you.
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1752411
Title:
bind9-host, avahi-daemon-check-dns.sh hang forever causes network
Confirmed the fix works on my affected system, after upgrade to
0.7-3.1ubuntu1.1 from bionic-proposed and a system reboot the boot works
(relatively) quickly as expected and doesn't get stuck.
Verified the file in place is the original from the package (and not one
modified by me).
Lastly to
** Tags removed: verification-needed-bionic
** Tags added: verification-done-bionic
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1752411
Title:
bind9-host, avahi-daemon-check-dns.sh hang forever
Hi Liam!
I also uninstalled avahi-daemon some months ago but I verified that the
problem came back when I installed it again, so I think you should still be
able to verify the fix if you want to.
Best regards
Laban
On Fri, Aug 31, 2018, 11:21 Liam wrote:
> I tried the numerous suggested
I tried the numerous suggested fixes and the all worked for me including
uninstalling "avahi-daemon". That last one resolved my problem so I
can't test the latest fix.
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
Reported upstream as https://gitlab.isc.org/isc-
projects/bind9/issues/520
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1752411
Title:
bind9-host, avahi-daemon-check-dns.sh hang forever causes
@Trent - since you had the most reproducible setup could you take a look
at verifying also the Bionic upload?
Also anyone else affected with the VPN cases please give it a try.
Finally, thanks Marc D. for adding all the Data.
If it is really hanging on that epoll like forever we might want to
Hello Liam, or anyone else affected,
Accepted avahi into bionic-proposed. The package will build now and be
available at https://launchpad.net/ubuntu/+source/avahi/0.7-3.1ubuntu1.1
in a few hours, and then in the -proposed repository.
Please help us by testing this new package. See
I found a source reference here:
https://github.com/fanf2/bind-9/blob/master/lib/isc/unix/socket.c#L4127
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1752411
Title:
bind9-host,
also attching a gdb full bt, which shows that epoll_wait is called it a
timeout value of "-1" (infinite) in thread #4.
** Attachment added: "gdb.txt"
https://bugs.launchpad.net/ubuntu/+source/openconnect/+bug/1752411/+attachment/5182603/+files/gdb.txt
--
You received this bug notification
in fact, there are two host commands running, both show
wchan=sigsuspend, perf shows nothing and strace shows commands is
suspended. new strace is attached.
** Attachment added: "strace.txt"
@Erich - infinite hangs are usually due to the kernel somewhere, while
suggesting dig was a good idea just to try I wonder if we would have to
find what "host" actually hangs on to be sure that "dig" in turn will
not some day block on just the same.
Can one of you affected when the "host" command
The suggested mitigation is in bionic-unapproved for the consideration
of the SRU team.
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1752411
Title:
bind9-host, avahi-daemon-check-dns.sh hang
** Merge proposal linked:
https://code.launchpad.net/~paelzer/ubuntu/+source/avahi/+git/avahi/+merge/353846
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1752411
Title:
bind9-host,
** Changed in: avahi (Ubuntu Cosmic)
Status: In Progress => Fix Released
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1752411
Title:
bind9-host, avahi-daemon-check-dns.sh hang forever
I've been trying to figure out how to test this with dig instead, and I think I
found something. If you have a normal /etc/resolv.conf then you see this:
===
$ dig -t soa local.; echo $?
; <<>> DiG 9.11.3-1ubuntu1.1-Ubuntu <<>> -t soa local.
;; global options: +cmd
;; Got answer:
;; WARNING:
> RE -W/w in `host`
, correct -- even with timeout set, it blocks forever (I tested this several
days ago in the dup'd ticket iirc)
> RE timeout
, good thoughts all - sure let's just stick with 5 seconds then
> RE logic true/false (@Trent)
, thanks yes! that'll do it; clarified now in my mind
> When host call fails (even with timeout), it returns "1" claiming
"dns_has_local()=true".
0 = true, 1 = false (you implied the opposite)
What may add confusion here is the grep -vq check is like an extra check
to make sure host didn't return 0 (success = we found .local) but then
say 'not
Thanks Trent for the extra verification.
Tests also look good so far, but currently since a lot got uploaded due to
feature freeze some tests take a while.
We should have that in cosmic soon and then can pick the same for Bionic.
Thanks also for your thoughts on a better long term solution.
--
I agree with the sentiment that 5 seconds feels too long, however as a
workaround I decided I would just copy the existing timeout. I certainly
would not want to make it longer since this is in the critical boot
path.
I would generally agree that in general a DNS request should fail faster
Hi fermulator,
I was wondering vice versa if 5 seconds would be too short actually.
Yes the good cases will return in sub-second, but it is the bad cases we want
to fix here.
Having a bit more time to recover if it can doesn't seem too bad to me.
On the question of host -W for waiting.
While it
We should also consider:
```
# CLEAN
fermulator@fermmy:~$ host -t soa local.
Host local. not found: 3(NXDOMAIN)
fermulator@fermmy:~$ echo $?
1
# BROKEN (host hangs)
fermulator@fermmy:~$ LC_ALL=C /usr/bin/timeout 1 host -t soa local. 2>&1
fermulator@fermmy:~$ echo $?
124
# timeout
PS: I've been running with a hacked /usr/lib/avahi/avahi-daemon-check-dns.sh
for a few days with this code:
```
OUT=`LC_ALL=C /usr/bin/timeout 2 host -t soa local. 2>&1`
```
, works like a charm
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed
Are we sure timeout of 5 seconds is appropriate? (it FEELS too long)
My intuition says that if a DNS query takes longer than 1 second it took too
long ...
However (consider also the "wait" (-W) parameter for the host command itself)
```
-W wait
Timeout: Wait for up to wait
I've had some minor cleanups on the changelog, but other than that I
think the most recent submission is good.
Also I found no issues in testing.
The code path it takes when the timeout triggers is that of a failing host
command (bad RC) which I think is just right. It will set things in a way
Request sponsorship of this upload for cosmic and then SRU to bionic
- New debdiff uploaded for both bionic and cosmic
- Fixed the SRU version for bionic
- Added a comment about the workaround to the script
- Updated bug description with SRU template
Tested patch working on bionic with my
** Patch added: "lp1752411-avahi-host-timeout-cosmic.patch"
https://bugs.launchpad.net/ubuntu/+source/avahi/+bug/1752411/+attachment/5178710/+files/lp1752411-avahi-host-timeout-cosmic.patch
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to
** Description changed:
+ [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 -
** Patch added: "lp1752411-avahi-host-timeout-bionic.patch"
https://bugs.launchpad.net/ubuntu/+source/avahi/+bug/1752411/+attachment/5178693/+files/lp1752411-avahi-host-timeout-bionic.patch
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to
** Patch added: "lp1752411-avahi-host-timeout-cosmic.patch"
https://bugs.launchpad.net/ubuntu/+source/avahi/+bug/1752411/+attachment/5178692/+files/lp1752411-avahi-host-timeout-cosmic.patch
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to
I agree strongswan/openconnect (and maybe more) are affected by the
symptom, while the bug lies in bind9-host/avahi packages at least
according to current debugging.
>From my experience I guess what would be great to get more traction on this is
>to get a shorter reproducer than setting up some
(this is currently in the "openconnect" path despite marked as "invalid"
against that package, bug was submitted originally to that project --
can we move to avahi?)
$ dpkg -S /usr/lib/avahi/avahi-daemon-check-dns.sh
avahi-daemon: /usr/lib/avahi/avahi-daemon-check-dns.sh
$ dpkg -s avahi-daemon
(btw; while we're fixing that script ... fix/change backtics to POSIX
compliant sub-shell'ing $() ?
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1752411
Title:
bind9-host,
Notes;
when "things are working", host does either:
while on VPN:
{{{
$ LC_ALL=C host -t soa local.
Host local. not found: 3(NXDOMAIN)
$ LC_ALL=C dig -t soa local.
; <<>> DiG 9.11.3-1ubuntu1.1-Ubuntu <<>> -t soa local.
;; global options: +cmd
;; Got answer:
;; WARNING: .local is reserved for
Definitely a duplicate. @paelzer was suggesting in bug 1786261 that dig
be used instead, do you guys know if that's a reasonable possibility?
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1752411
** Also affects: strongswan (Ubuntu)
Importance: Undecided
Status: New
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1752411
Title:
bind9-host, avahi-daemon-check-dns.sh hang forever
FYI bug 1786261 could be another symptoms of this, reporters there will
take a look and might add another affected package to this.
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1752411
Title:
Unsubscribing sponsors for now, awaiting the fixes Robie commented
about.
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1752411
Title:
bind9-host, avahi-daemon-check-dns.sh hang forever causes
Thank you for working on this. I agree with your approach. The debdiff
looks good.
I think that though it's clear that the bug is in the host command,
given that we haven't been able to figure out the fix (I spent some time
on it too), it's reasonable to add the timeout command as in your patch
** Tags added: server-next
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1752411
Title:
bind9-host, avahi-daemon-check-dns.sh hang forever causes network
connections to get stuck
To manage
Hoping to get attention to this again. Since 18.04.1 is out now, more
and more users are likely to hit this issue as more users will be
upgrading. This issue applies equally to desktop and server scenarios.
I would like to get lp1752411-avahi-host-timeout.diff sponsored for
upload please
--
You
I've run into this problem as well after upgrading my Ubuntu 17.10
installation (upgraded from 17.04) to Ubuntu 18.04 last week. My VPN
script calling "openconnect" hang and no packets got forwarded. My first
workaround was however to connect using nm-applet instead which worked
fine (available
I was hit by this exact problem (update to Ubuntu 18.04), the host
command would hang and adding timeout works around the issue. Please
put the temporary fix in so others are not affected.
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to
Sponsors: Can we get this debdiff uploaded now? We've had a few more
reports and I'd like to get this workaround in place.
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1752411
Title:
bind9-host,
** Changed in: avahi (Debian)
Status: Unknown => New
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1752411
Title:
bind9-host, avahi-daemon-check-dns.sh hang forever causes network
It should probably also check for 127.0.0.53, not just 127.0.0.1
** Also affects: avahi (Debian) via
https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=898038
Importance: Unknown
Status: Unknown
--
You received this bug notification because you are a member of Ubuntu
Server, which
It should probably also check for 127.0.0.53, not just 127.0.0.1
** Also affects: avahi (Debian) via
https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=898038
Importance: Unknown
Status: Unknown
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is
Please also see https://bugs.debian.org/898038
** Bug watch added: Debian Bug tracker #898038
https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=898038
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
I think that time is past, we were in beta freeze in the past week, and
are in final freeze now. Unless there is a clear test case showing under
which conditions this happens and how widespread it is, it's probably
best to start thinking in SRU terms.
It looks like a safe change, but since I
I think that time is past, we were in beta freeze in the past week, and
are in final freeze now. Unless there is a clear test case showing under
which conditions this happens and how widespread it is, it's probably
best to start thinking in SRU terms.
It looks like a safe change, but since I
I'd still like to get the upload debdiff for 'timeout' that I prepared
uploaded. Even if we manage to debug the bind9-host issue, it will
still be useful to have the timeout command there as a backup. Not long
before we run out of time for bionic release.
I am actively looking at the bind9-host
Some troubleshooting I did with Trent today showed:
a) the "host -t soa local." call triggered a query to 127.0.0.53 as
expected, network-wise, which got a response right away
b) we snapshotted ip route and ip addr just before the host call, and
saw that the interface responsible for the default
Some troubleshooting I did with Trent today showed:
a) the "host -t soa local." call triggered a query to 127.0.0.53 as
expected, network-wise, which got a response right away
b) we snapshotted ip route and ip addr just before the host call, and
saw that the interface responsible for the default
There is a new bind9 upload to bionic-proposed (9.11.3+dfsg-1ubuntu1)
Tested with this version and 'host' is still hanging. So this fix is
still required.
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
** Patch added: "lp1752411-avahi-host-timeout.diff"
https://bugs.launchpad.net/ubuntu/+source/avahi/+bug/1752411/+attachment/5117372/+files/lp1752411-avahi-host-timeout.diff
** Changed in: avahi (Ubuntu)
Assignee: (unassigned) => Trent Lloyd (lathiat)
--
You received this bug
With host -d I simply get
> Trying "local"
When it works normally I get;
Trying "local"
Host local. not found: 3(NXDOMAIN)
Received 98 bytes from 10.48.134.6#53 in 1 ms
Received 98 bytes from 10.48.134.6#53 in 1 ms
The system I am hitting this issue on is an upgraded system (rather than
a
My system was upgraded, so maybe a leftover. So I uninstalled ifupdown and
upstart. Problem still persists :-(
I will attach a ltrace -Sf host output - maybe it helps...
** Attachment added: "output of "trace -Sf host -d -v -w5 -t soa local.""
Can you try running "host" with -d in the scenario where it is hanging?
Also, a fresh bionic system shouldn't run the ifup-down scripts, since
it uses netplan. Unless openconnect or one of its dependencies pull it
in explicitly. A quick apt-cache rdepends didn't see it.
--
You received this bug
Can you try running "host" with -d in the scenario where it is hanging?
Also, a fresh bionic system shouldn't run the ifup-down scripts, since
it uses netplan. Unless openconnect or one of its dependencies pull it
in explicitly. A quick apt-cache rdepends didn't see it.
--
You received this bug
I did some testing using strace and looking at backtraces of why "host"
is stuck, and it's not immediately clear to me why it's getting stuck.
Will need to look more in depth into it tracing it's actual execution -
it's multi threaded and using poll so not super straight forward from
the trace for
67 matches
Mail list logo
