Ubuntu
ntp package

[SRU] ntp wont run on IPv6 only host unless lo has 127.0.0.1 address

Bug #715152 reported by dnmvisser on 2011-02-08

This bug affects 1 person

	Status	Importance	Assigned to
NTP	Unknown	Unknown	auto-bugs.ntp.org #397
ntp (Ubuntu)	Fix Released	Medium	Unassigned
Lucid	Fix Released	Medium	James Page
Maverick	Fix Released	Medium	James Page

Bug Description

IMPACT:
Ubuntu instances running in pure IPv6 environments (including 127.0.0.1) cannot run ntp as it dies when scanning for available interfaces.

FIX:
IPv6/IPv4 scanning has been improved in later releases; patch created from later release to resolve this issue (directly attached to this bug report).

TEST CASE:
1) On a working system with IPv4 & IPV6:
sudo apt-get install ntp
netstat -a | grep ntp
Ports should be mapped for both udp and udp6

2) Disable IPv4
Add the following line to /etc/network/interfaces to disable 127.0.0.1:

iface lo inet loopback
post-up ip addr del 127.0.0.1/8 dev lo

And remove the line(s) relating to IPv4 configuration of other network interfaces (looks like iface eth* inet dhcp|static*)

3) Reboot server

4) NTP service will not be running; messages in /var/log/syslog relating to startup of daemon; however no daemon running
ps -aef | grep ntp
netstat -a | grep ntp

IMPACT:
Patch is relatively minimal and comes from later release.

>>>> ORIGINAL BUG REPORT <<<<<

Binary package hint: ntp

Lucid host running IPv6 only, with network configured like this:

root@svn:~# ifconfig
eth0 Link encap:Ethernet HWaddr 00:50:56:86:00:15
          inet6 addr: 2001:610:148:dead:cafe:babe:0:29/64 Scope:Global
          inet6 addr: fe80::250:56ff:fe86:15/64 Scope:Link
          UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
          RX packets:3164 errors:0 dropped:0 overruns:0 frame:0
          TX packets:2075 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000
          RX bytes:1516490 (1.5 MB) TX bytes:387497 (387.4 KB)

lo Link encap:Local Loopback
          inet6 addr: ::1/128 Scope:Host
          UP LOOPBACK RUNNING MTU:16436 Metric:1
          RX packets:0 errors:0 dropped:0 overruns:0 frame:0
          TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:0
          RX bytes:0 (0.0 B) TX bytes:0 (0.0 B)

Now ntpd wont start:

./../libisc/ifiter_ioctl.c:567: REQUIRE(iter->pos < (unsigned int) iter->ifc.ifc_len) failed.
Aborted

After adding an IPv4 loopback address to lo interface:

ip addr add 127.0.0.1 dev lo

ntpd is able to start.

According to http://bugs.ntp.org/show_bug.cgi?id=397 this might be related to libisc.

See original description

Tags:

Related branches

lp:~james-page/ubuntu/lucid/ntp/fix-715152

Merged into lp:ubuntu/lucid/ntp

Dave Walker (community): Approve on 2011-04-15

lp:~james-page/ubuntu/maverick/ntp/fix-715152

Merged into lp:ubuntu/maverick/ntp

Dave Walker (community): Approve on 2011-04-15

lp:ubuntu/lucid-proposed/ntp

lp:ubuntu/maverick-proposed/ntp

Revision history for this message

James Page (james-page) wrote on 2011-02-10:

Thank you for taking the time to report this bug and helping to make Ubuntu better. Please execute the following command, as it will automatically gather debugging information, in a terminal:

apport-collect 715152

When reporting bugs in the future please use apport by using 'ubuntu-bug' and the name of the package affected. You can learn more about this functionality at https://wiki.ubuntu.com/ReportingBugs.

Changed in ntp (Ubuntu):
status:	New → Incomplete

Revision history for this message

James Page (james-page) wrote on 2011-02-10:

Confirmed on a Lucid image; however the only way I was able to remove the 127.0.0.1 address from lo was to use the following command:

ip addr del 127.0.0.1/32 dev lo

ntp then fails to start. As stated in the original bug report, adding the address back to the interface fixes the issue.

How where you able to permanently disable the ipv4 address of the lo device?

Changed in ntp (Ubuntu):
status:	Incomplete → Confirmed
status:	Confirmed → Incomplete

Revision history for this message

dnmvisser (dnmvisser) wrote on 2011-02-11:

Ehm, well exactly like you did, I manually removed it, although it's done automatically from interfaces with 'pre-up'.

Revision history for this message

James Page (james-page) wrote on 2011-02-11: Re: [Bug 715152] Re: ntp wont run on IPv6 only host unless lo has 127.0.0.1 address

On Fri, 2011-02-11 at 12:04 +0000, /usr/local/dick wrote:
> Ehm, well exactly like you did, I manually removed it, although it's
> done automatically from interfaces with 'pre-up'.
>

Guessed that might be the case. I'll get this configured in my test
rig.

--
James Page
Software Engineer, Ubuntu Server Team

Revision history for this message

James Page (james-page) wrote on 2011-04-06: Re: ntp wont run on IPv6 only host unless lo has 127.0.0.1 address

I've done some further investigation into this issue.

maverick exhibits the same symptoms as lucid; i.e. if the 127.0.0.1 address is not bound ntp exits on startup.

natty appears to deal with this scenario OK; removing 127.0.0.1 in a post-up step does not impact ntp.

Looks like this issue was fixed in the transition from v4.2.4 to v4.2.6 (possibly as part of 4.2.5p219)

Revision history for this message

James Page (james-page) wrote on 2011-04-06:

ntp-error.log Edit (94 bytes, text/plain)

Output from running ntpd -n on a maverick host with no ipv4 enabled including localhost.

This code has changed maverick->natty to deal with ipv6 better (by the looks of it).

Revision history for this message

James Page (james-page) wrote on 2011-04-06:

So this issue related to upstrem bug http://bugs.ntp.org/show_bug.cgi?id=397

James Page (james-page) on 2011-04-06

Changed in ntp (Ubuntu):
status:	Incomplete → Fix Released
importance:	Undecided → Medium

Revision history for this message

James Page (james-page) wrote on 2011-04-11:

fix-noipv4.patch Edit (2.3 KiB, text/plain)

Minimal patch to fix running with no IPv4

description:

updated

James Page (james-page) on 2011-04-11

description:

updated

James Page (james-page) on 2011-04-15

summary:	- ntp wont run on IPv6 only host unless lo has 127.0.0.1 address + [SRU] ntp wont run on IPv6 only host unless lo has 127.0.0.1 address
Changed in ntp (Ubuntu Lucid):
assignee:	nobody → James Page (james-page)
Changed in ntp (Ubuntu Maverick):
assignee:	nobody → James Page (james-page)
Changed in ntp (Ubuntu Lucid):
status:	New → In Progress
Changed in ntp (Ubuntu Maverick):
status:	New → In Progress
Changed in ntp (Ubuntu Lucid):
importance:	Undecided → Medium
Changed in ntp (Ubuntu Maverick):
importance:	Undecided → Medium

Revision history for this message

Clint Byrum (clint-fewbar) wrote on 2011-04-19:

In reviewing this diff for lucid-proposed:

http://launchpadlibrarian.net/69484036/ntp_1%3A4.2.4p8%2Bdfsg-1ubuntu2_1%3A4.2.4p8%2Bdfsg-1ubuntu2.1.diff.gz

and the other similar diff for maverick-proposed;

It took me a while to see why it was "Forwarded: not-needed". The Origin header of the patch would usually tell me, but that just shows me where this file was added to Natty in a merge. Its fairly difficult to tell that this patch is just the differences for this file between 4.2.4 and 4.2.6, which the linked upstream bug report tells me is the revision in which the bug has been fixed.

I don't see where our time will be used well by tracking that down much further, but I wanted to make it clear that this is not very straight forward.

Given that, we need to see wide spread testing of this before it leaves -proposed.

APPROVED: the package version 1:4.2.4p8+dfsg-1ubuntu2.1 uploaded to lucid-proposed should be accepted.
APPROVED: the package version 1:4.2.4p8+dfsg-1ubuntu6.1 uploaded to maverick-proposed should be accepted.

Revision history for this message

Martin Pitt (pitti) wrote on 2011-04-19: Please test proposed package

#10

Accepted ntp into lucid-proposed, the package will build now and be available in a few hours. Please test and give feedback here. See https://wiki.ubuntu.com/Testing/EnableProposed for documentation how to enable and use -proposed. Thank you in advance!

Changed in ntp (Ubuntu Lucid):
status:	In Progress → Fix Committed
tags:	added: verification-needed

Revision history for this message

Martin Pitt (pitti) wrote on 2011-04-19:

#11

Accepted ntp into maverick-proposed, the package will build now and be available in a few hours. Please test and give feedback here. See https://wiki.ubuntu.com/Testing/EnableProposed for documentation how to enable and use -proposed. Thank you in advance!

Changed in ntp (Ubuntu Maverick):
status:	In Progress → Fix Committed

Revision history for this message

dnmvisser (dnmvisser) wrote on 2011-04-20:

#12

Just installed "ntp 1:4.2.4p8+dfsg-1ubuntu2.1" from lucid-proposed, and that seems to do the trick.
Thanks for fixing this guys!

And well in advance of 6-6, IPv6 day ;-)

Martin Pitt (pitti) on 2011-04-20

tags:

added: verification-done
removed: verification-needed

Revision history for this message

Launchpad Janitor (janitor) wrote on 2011-04-26:

#13

This bug was fixed in the package ntp - 1:4.2.4p8+dfsg-1ubuntu2.1

---------------
ntp (1:4.2.4p8+dfsg-1ubuntu2.1) lucid-proposed; urgency=low

* debian/patches/fix-noipv4.patch: support running in IPv6 only
environments (LP: #715152).
-- James Page <email address hidden> Mon, 11 Apr 2011 11:27:28 +0100

Changed in ntp (Ubuntu Lucid):
status:	Fix Committed → Fix Released

Revision history for this message

Launchpad Janitor (janitor) wrote on 2011-04-26:

#14

This bug was fixed in the package ntp - 1:4.2.4p8+dfsg-1ubuntu6.1

---------------
ntp (1:4.2.4p8+dfsg-1ubuntu6.1) maverick-proposed; urgency=low

* debian/patches/fix-noipv4.patch: support running in IPv6 only
environments (LP: #715152).
-- James Page <email address hidden> Wed, 06 Apr 2011 11:19:26 +0100