FS#31775 - [linux] 3.6-1 ntpq request timed out

Attached to Project: Arch Linux
Opened by H.pferd (stosch) - Tuesday, 02 October 2012, 22:31 GMT
Last edited by Tom Gundersen (tomegun) - Thursday, 01 November 2012, 23:45 GMT
Task Type Bug Report
Category Packages: Core
Status Closed
Assigned To Tobias Powalowski (tpowa)
Thomas Bächler (brain0)
Architecture x86_64
Severity Low
Priority Normal
Reported Version
Due in Version Undecided
Due Date Undecided
Percent Complete 100%
Votes 0
Private No

Details

Description:
The command "ntpq -p" does not work any more with current testing kernel (3.6.0-1).
The error message is:

localhost.localdomain: timed out, nothing received
***Request timed out

According to systemctl ntp is still running:

ntpd.service loaded active running Network Time Service

Downgrading to stable kernel 3.5.4 solves the problem.
I have a fully updated system with testing-repo activated.

If you need any more information/logs please tell me.

Additional info:
ntp 4.2.6.p5-11
linux 3.6.0-1
This task depends upon

Closed by  Tom Gundersen (tomegun)
Thursday, 01 November 2012, 23:45 GMT
Reason for closing:  Fixed
Additional comments about closing:  3.6.4-1
Comment by Dave Reisner (falconindy) - Tuesday, 02 October 2012, 23:11 GMT
"ntpq -4 -p" works. A simple strace would have shown you ntpq connecting on the ipv6 loopback instead of the ipv4.
Comment by H.pferd (stosch) - Wednesday, 03 October 2012, 00:16 GMT
Well, if this is expected behaviour the report should be closed.
But if I understand you right you have the same experience as me and I just checked it on another system and the old syntax (ntpq -p) worked without problems (same kernelversion etc), what makes me wonder why..
Comment by Dave Reisner (falconindy) - Wednesday, 03 October 2012, 00:21 GMT
I'm only pointing out why it fails. Whether or not this needs to be "fixed" and where is a different issue.
Comment by H.pferd (stosch) - Wednesday, 03 October 2012, 19:03 GMT
So, here are the results of my investigations:
It seems to be related somehow to avahi/cups because if I run
systemctl restart avahi-daemon.service ntpd
or disable cups (systemctl disable...) I can restore the old behaviour (ntpq -p working).
I cannot get any clues from that.

@falconindy:
Is it correct that you have the same "problem"? Does my wannabe-workaround help you, too?
I ask because I still cannot reproduce the error on my second machine..
Comment by Dave Reisner (falconindy) - Wednesday, 03 October 2012, 19:11 GMT
I don't consider it relevant because I'm not using avahi.
Comment by H.pferd (stosch) - Monday, 08 October 2012, 16:55 GMT
Just a short status report:
No changes with linux 3.6.1-1.
As it is probably an upstream problem: Should I report it upstream?
Comment by H.pferd (stosch) - Saturday, 13 October 2012, 17:21 GMT
FYI I just reported it to the kernel bugtracker:
https://bugzilla.kernel.org/show_bug.cgi?id=48741
Comment by H.pferd (stosch) - Monday, 15 October 2012, 13:11 GMT
Once again FYI: Here is a discussion regarding the problem on the netdev-mailing list:
http://marc.info/?l=linux-netdev&m=135030432513727&w=2

@falconindy:
Can you confirm, that /proc/net/if_inet6 does not contain the lo interface?
Comment by Dave Reisner (falconindy) - Monday, 15 October 2012, 13:17 GMT
$ cat /proc/net/if_inet6
fe8000000000000064fe82fffe0ba0d4 03 40 20 80 tap0
fe80000000000000beaec5fffe0bfcc8 02 40 20 80 eth0
Comment by H.pferd (stosch) - Monday, 15 October 2012, 14:55 GMT
Thanks.
Comment by H.pferd (stosch) - Tuesday, 16 October 2012, 19:25 GMT
A patch has been posted, to resolve the issue:
http://marc.info/?l=linux-netdev&m=135040905619911&w=2
So, this can be closed for me. Thank you anyway.

Loading...