| View previous topic :: View next topic |
| Author |
Message |
hanj
Veteran
Joined: 19 Aug 2003
Posts: 1500
|
 Posted: Mon Mar 05, 2007 3:27 pm Post subject: New NTP error after upgrade |
 |
|
Hello
I just upgraded ntp to 4.2.4-r1 and I'm starting to see some errors in the logs.
| Code: | Mar 5 07:20:09 comp ntpd[7181]: kernel time sync error 0001
Mar 5 07:54:18 comp ntpd[7181]: kernel time sync error 0001 |
Can anyone shed some light on this? Let me know if I can provide any additional information.
Here are my USE flags:
| Code: | | net-misc/ntp-4.2.4-r1 USE="caps ssl -debug -ipv6 -openntpd -parse-clocks (-selinux)" |
Thanks!
hanji
_________________
Server Admin Blog - Uno-Code.com |
|
| Back to top |
|
 |
wynn
Advocate
Joined: 01 Apr 2005
Posts: 2421
Location: UK
|
| Posted: Mon Mar 05, 2007 3:58 pm Post subject: |
|
|
It is a bug in 4.2.4 (apparently) see https://ntp.isc.org/bugs/show_bug.cgi?id=780. The last message is 26 February but, without reading and understanding the exchanges, it it not clear whether it has been resolved.
_________________
The avatar is jorma, a "duck" from "Elephants Dream": the film and all the production materials have been made available under a Creative Commons Attribution 2.5 License, see orange.blender.org for details. |
|
| Back to top |
|
|
hanj
Veteran
Joined: 19 Aug 2003
Posts: 1500
|
|
| Back to top |
|
|
deoren
Apprentice
Joined: 23 Dec 2004
Posts: 227
Location: USA
|
| Posted: Wed Mar 07, 2007 6:57 pm Post subject: |
|
|
Thanks wynn for directing me to this post. Also thanks for your sig because now I've added that movie to my todo list. 
In case anyone doesn't know, one possible workaround is to downgrade to net-misc/ntp-4.2.2_p3 until the problem is worked out upstream.
This can be accomplished by adding the following:
| Code: | # http://forums.gentoo.org/viewtopic-t-544792.html
=net-misc/ntp-4.2.4-r1 |
to your /etc/portage/package.mask file and running
This should show you that net-misc/ntp-4.2.2_p3 will be installed.
Hope that helps someone. |
|
| Back to top |
|
|
overkll
Veteran
Joined: 21 Sep 2004
Posts: 1249
Location: Austin, Texas
|
| Posted: Wed Mar 07, 2007 7:38 pm Post subject: |
|
|
| Funny, I'm not experiencing ANY errors with ntp-4.2.4-r1. |
|
| Back to top |
|
|
AllanS
n00b
Joined: 19 Oct 2006
Posts: 48
|
| Posted: Fri Mar 09, 2007 2:20 pm Post subject: |
|
|
I just upgraded and now ntpd will not start:
| Code: | ntpd[5607]: ntpd 4.2.4@1.1437-o Thu Mar 8 21:55:16 UTC 2007 (1)
ntpd[5608]: precision = 1.000 usec
ntpd[5608]: Listening on interface #0 wildcard, 0.0.0.0#123 Disabled
ntpd[5608]: Listening on interface #1 eth0, 192.169.1.42#123 Enabled
ntpd[5608]: Listening on interface #2 lo, 127.0.0.1#123 Enabled
ntpd[5608]: kernel time sync status 0040
rc-scripts: Failed to start ntpd
|
| Code: | uname -a
Linux galactica 2.6.19-gentoo-r5 #1 PREEMPT Thu Mar 1 15:50:51 EST 2007 i686 AMD Athlon(TM) XP1600+ AuthenticAMD GNU/Linux |
This is after a "emerge --update --ask --deep --newuse world" and "revdep-rebuild". etc-update was done as well. |
|
| Back to top |
|
|
AllanS
n00b
Joined: 19 Oct 2006
Posts: 48
|
| Posted: Fri Mar 09, 2007 3:05 pm Post subject: |
|
|
I just downgraded my ntpd and it still does not work:
| Code: | ntpd[19270]: ntpd 4.2.2p3@1.1577-o Fri Mar 9 15:01:22 UTC 2007 (1)
ntpd[19271]: precision = 1.000 usec
ntpd[19271]: Listening on interface wildcard, 0.0.0.0#123 Disabled
ntpd[19271]: Listening on interface eth0, 192.169.1.42#123 Enabled
ntpd[19271]: Listening on interface lo, 127.0.0.1#123 Enabled
ntpd[19271]: kernel time sync status 0040
rc-scripts: Failed to start ntpd |
Any other ideas? |
|
| Back to top |
|
|
overkll
Veteran
Joined: 21 Sep 2004
Posts: 1249
Location: Austin, Texas
|
| Posted: Fri Mar 09, 2007 4:03 pm Post subject: |
|
|
Allan, what's your /etc/ntp.conf and /etc/conf.d/ntpd contain?
Are you using dhcp to get an IP address? |
|
| Back to top |
|
|
AllanS
n00b
Joined: 19 Oct 2006
Posts: 48
|
| Posted: Fri Mar 09, 2007 5:21 pm Post subject: |
|
|
I use a static IP through a DSL router. The router gets its IP address from my ISP. Network connectivity, name resolution, etc., all work fine. I can ping my NTP server and using the "ntp-client" during startup works fine as well (to the same server). All this worked before the upgrade.
/etc/conf.d/ntpd :
| Code: | # /etc/conf.d/ntpd
# Options to pass to the ntpd process
# Most people should leave this line alone ...
# however, if you know what you're doing, feel free to tweak
NTPD_OPTS="-u ntp:ntp" |
Non-comment sections of /etc/ntp.conf:
| Code: | server timex.peachnet.edu
driftfile /var/lib/ntp/ntp.drift
restrict default nomodify nopeer
restrict 127.0.0.1
restrict 192.169.0.0 mask 255.255.255.0 nomodify nopeer notrap |
I am happy to be a guinea pig to figure this out, but I'm not sure which direction I should go in. What other information would be helpful? I am back using ntpd 4.2.4 but since the older version did not work, I am wondering if there is another library or package that is interfering with this.
I also recently emerged cdrkit which required cmake and libcap (libcap is now version 1.10-r9, but I think it was already at that version, so maybe a IUSE change?).
I checked for other applications binding to port 123 (ntp) and found none. |
|
| Back to top |
|
|
doctorofcomputers
n00b
Joined: 09 Sep 2006
Posts: 2
|
| Posted: Fri Mar 09, 2007 8:12 pm Post subject: |
|
|
*edit*
Last edited by doctorofcomputers on Sat Mar 10, 2007 7:00 pm; edited 1 time in total |
|
| Back to top |
|
|
AllanS
n00b
Joined: 19 Oct 2006
Posts: 48
|
| Posted: Fri Mar 09, 2007 9:06 pm Post subject: |
|
|
Well, in the spirit of "What the heck", I tried it. Full removal and then full install.
It didn't work for me. |
|
| Back to top |
|
|
wynn
Advocate
Joined: 01 Apr 2005
Posts: 2421
Location: UK
|
| Posted: Fri Mar 09, 2007 9:18 pm Post subject: |
|
|
ntpd will fail to start if the difference between the time on your box and the server time is over 1000 seconds, could this be the case?
You should be able to see messages in /var/log/messages showing what went on.
ntpdate (run by ntp-client) will timeout if it can't get a server response in 1 second (default) and may not, therefore, have corrected the BIOS time.
_________________
The avatar is jorma, a "duck" from "Elephants Dream": the film and all the production materials have been made available under a Creative Commons Attribution 2.5 License, see orange.blender.org for details. |
|
| Back to top |
|
|
AllanS
n00b
Joined: 19 Oct 2006
Posts: 48
|
| Posted: Fri Mar 09, 2007 10:02 pm Post subject: |
|
|
I do not believe that is the case. The current date on my box is very close to other computers and matches the wall clock (within a minute or so). It is well within the range of 1000 seconds.
I just ran ntpdate manually against my normal server and it replied that I was 0.122 seconds off. This shows that ntpdate works and so does the server.
The time is correct (via ntp-client), but ntpd will not start.
This seems to be specific to ntpd, but not just this version but the last as well. Since the last version used to work I am thinking that it is a recent library or supporting function that ntpd uses. A quick "ldd -v /usr/sbin/ntpd" shows:
| Code: | linux-gate.so.1 => (0xffffe000)
libm.so.6 => /lib/libm.so.6 (0xb7f16000)
libcrypto.so.0.9.8 => /usr/lib/libcrypto.so.0.9.8 (0xb7dde000
libc.so.6 => /lib/libc.so.6 (0xb7cb7000)
/lib/ld-linux.so.2 (0xb7f60000)
libdl.so.2 => /lib/libdl.so.2 (0xb7cb3000)
Version information:
/usr/sbin/ntpd:
libm.so.6 (GLIBC_2.0) => /lib/libm.so.6
libc.so.6 (GLIBC_2.1.3) => /lib/libc.so.6
libc.so.6 (GLIBC_2.3) => /lib/libc.so.6
libc.so.6 (GLIBC_2.1) => /lib/libc.so.6
libc.so.6 (GLIBC_2.3.4) => /lib/libc.so.6
libc.so.6 (GLIBC_2.0) => /lib/libc.so.6
/lib/libm.so.6:
ld-linux.so.2 (GLIBC_PRIVATE) => /lib/ld-linux.so.2
libc.so.6 (GLIBC_2.1.3) => /lib/libc.so.6
libc.so.6 (GLIBC_2.0) => /lib/libc.so.6
/usr/lib/libcrypto.so.0.9.8:
libdl.so.2 (GLIBC_2.1) => /lib/libdl.so.2
libdl.so.2 (GLIBC_2.0) => /lib/libdl.so.2
libc.so.6 (GLIBC_2.1.3) => /lib/libc.so.6
libc.so.6 (GLIBC_2.1) => /lib/libc.so.6
libc.so.6 (GLIBC_2.3) => /lib/libc.so.6
libc.so.6 (GLIBC_2.0) => /lib/libc.so.6
/lib/libc.so.6:
ld-linux.so.2 (GLIBC_2.1) => /lib/ld-linux.so.2
ld-linux.so.2 (GLIBC_2.3) => /lib/ld-linux.so.2
ld-linux.so.2 (GLIBC_PRIVATE) => /lib/ld-linux.so.2
/lib/libdl.so.2:
libc.so.6 (GLIBC_2.1.3) => /lib/libc.so.6
libc.so.6 (GLIBC_2.1) => /lib/libc.so.6
libc.so.6 (GLIBC_2.0) => /lib/libc.so.6
libc.so.6 (GLIBC_PRIVATE) => /lib/libc.so.6
ld-linux.so.2 (GLIBC_PRIVATE) => /lib/ld-linux.so.2 |
All of these were compiled around 22-Feb-2007, but the ntpd was working after that. Only libcap was comiled on 07-Mar-2007, so that is why I suspect it. |
|
| Back to top |
|
|
overkll
Veteran
Joined: 21 Sep 2004
Posts: 1249
Location: Austin, Texas
|
| Posted: Fri Mar 09, 2007 10:46 pm Post subject: |
|
|
| Have you tried fiddling with /etc/ntpd.conf? Try commenting out everything except for the server and the driftfile location, save and restart. |
|
| Back to top |
|
|
AllanS
n00b
Joined: 19 Oct 2006
Posts: 48
|
| Posted: Sat Mar 10, 2007 1:38 am Post subject: |
|
|
No joy. I haven't touch that file in months. So I am not surprised that did not fix it.
I looked into the kernel code (ntp.c) and see that the status listed above (in my log: 0x0040) is STA_UNSYNC - this is the default error status. I did recently upgrade my kernel from 2.6.18-gentoo-r6 to 2.6.19-gentoo-r5 and I am beginning to think it is associated with that. In the older kernel the default error status was 0x0000. In going to the new kernel, all the ntp-related functions in time.c and timer.c were moved to a new ntp.c and that is where the default status is set.
Can someone in the kernel team look there and see if I am interpreting this correctly? It may be that ntpd just needs to ignore that status at that time of the process and just continue on. |
|
| Back to top |
|
|
wynn
Advocate
Joined: 01 Apr 2005
Posts: 2421
Location: UK
|
| Posted: Sat Mar 10, 2007 12:04 pm Post subject: |
|
|
ldd -v /usr/sbin/ntpd showed exactly the same output as you have â libpcap isn't there though: ntpd opens an ordinary socket pair to send and receive messages on UDP port 123, not like tcpdump which wants to get everything.
What do you find in /var/log/messages? ntpd should say why it is stopping and this will give us a clue as to where to concentrate our efforts.
Are you running it in a chroot jail and it can't get at its ntp.conf?
Does your /etc/ntp.drift look reasonable? This is what it has here | Code: | # cat /var/lib/ntp/ntp.drift
46.298 |
for a working 4.2.4-r1.
http://www.ee.udel.edu/~mills/ntp/html/debug.html says you can run ntpd in debugging mode with the "-d" flag â but it doesn't work.
[Edit] It needs to be emerge'd with the "debug" USE flag. You should get something like this | Code: | ntpd -d
ntpd 4.2.4@1.1437-o Sat Mar 10 16:00:06 UTC 2007 (1)
addto_syslog: precision = 1.000 usec
addto_syslog: ntp_io: estimated max descriptors: 1024, initial socket boundary:
16
addto_syslog: Listening on interface #0 wildcard, 0.0.0.0#123 Disabled
addto_syslog: no IPv6 interfaces found
addto_syslog: no IPv6 interfaces found
addto_syslog: Listening on interface #1 lo, 127.0.0.1#123 Enabled
addto_syslog: Listening on interface #2 eth0, 192.168.1.37#123 Enabled
addto_syslog: Listening on interface #3 ath0, 192.168.0.100#123 Enabled
local_clock: time 0 offset 0.000000 freq 0.000 state 0
addto_syslog: kernel time sync status 0040
...
newpeer: 192.168.1.37->194.106.50.10 mode 3 vers 4 poll 6 10 flags 0x1 0x1 ttl 0 key 00000000
... |
and so on. "newpeer" should have the IP address of your server and the "addto_syslog" are the entries which go into /var/log/messages â unless your syslog.conf/syslog-ng.conf arranges for them to go somewhere else.
_________________
The avatar is jorma, a "duck" from "Elephants Dream": the film and all the production materials have been made available under a Creative Commons Attribution 2.5 License, see orange.blender.org for details. |
|
| Back to top |
|
|
wynn
Advocate
Joined: 01 Apr 2005
Posts: 2421
Location: UK
|
| Posted: Sat Mar 10, 2007 4:26 pm Post subject: |
|
|
| AllanS wrote: | No joy. I haven't touch that file in months. So I am not surprised that did not fix it.
I looked into the kernel code (ntp.c) and see that the status listed above (in my log: 0x0040) is STA_UNSYNC - this is the default error status. I did recently upgrade my kernel from 2.6.18-gentoo-r6 to 2.6.19-gentoo-r5 and I am beginning to think it is associated with that. In the older kernel the default error status was 0x0000. In going to the new kernel, all the ntp-related functions in time.c and timer.c were moved to a new ntp.c and that is where the default status is set.
Can someone in the kernel team look there and see if I am interpreting this correctly? It may be that ntpd just needs to ignore that status at that time of the process and just continue on. |
It will always start as unsynchronized, ntpd has to do quite a bit of work (file:///usr/share/doc/ntp-4.2.4-r1/html/debug.html) before setting the clock | Quote: | | Note that the daemon requires at least four packet exchanges when first started in any case. This is required in order for the mitigation algorithms to insure valid and accurate measurements and defend against network delay spikes and accidental or malicious errors induced by the servers selected in the configuration file. It normally takes less than four minutes to set the clock when first started, but this can be reduced to less than ten seconds with the iburst configuration option. |
The startup messages in /var/log/messages show this | Code: | Mar 10 16:07:12 lightfoot ntpd[31372]: kernel time sync status 0040
Mar 10 16:07:12 lightfoot ntpd[31372]: frequency initialized 52.003 PPM from /var/lib/ntp/ntp.drift
Mar 10 16:11:28 lightfoot ntpd[31372]: synchronized to 212.13.197.135, stratum 2
Mar 10 16:11:28 lightfoot ntpd[31372]: kernel time sync status change 0001 |
starting with your STA_UNSYNC and then changing after it has found a server it can trust.
_________________
The avatar is jorma, a "duck" from "Elephants Dream": the film and all the production materials have been made available under a Creative Commons Attribution 2.5 License, see orange.blender.org for details. |
|
| Back to top |
|
|
overkll
Veteran
Joined: 21 Sep 2004
Posts: 1249
Location: Austin, Texas
|
| Posted: Sat Mar 10, 2007 4:48 pm Post subject: |
|
|
| Maybe it is kernel related. I'm using ck-sources and have no errors. |
|
| Back to top |
|
|
wynn
Advocate
Joined: 01 Apr 2005
Posts: 2421
Location: UK
|
| Posted: Sat Mar 10, 2007 5:58 pm Post subject: |
|
|
Well, ntpd works without trouble on 2.6.18-gentoo-r3, 2.6.19-gentoo-r5 and 2.6.10-gentoo-r2 (the late compilation timestamp "Sat Mar 10 16:00:06 UTC 2007" is because I emerge'd it again with debugging).
2.6.18-gentoo-r3 | Code: | Mar 10 17:21:08 lightfoot ntpd[6902]: ntpd 4.2.4@1.1437-o Sat Mar 10 16:00:06 UTC 2007 (1)
Mar 10 17:21:08 lightfoot ntpd[6903]: precision = 1.000 usec
Mar 10 17:21:08 lightfoot ntpd[6903]: Listening on interface #0 wildcard, 0.0.0.0#123 Disabled
Mar 10 17:21:08 lightfoot ntpd[6903]: Listening on interface #1 lo, 127.0.0.1#123 Enabled
Mar 10 17:21:08 lightfoot ntpd[6903]: kernel time sync status 0040
Mar 10 17:21:08 lightfoot ntpd[6903]: frequency initialized 57.084 PPM from /var/lib/ntp/ntp.drift
Mar 10 17:21:09 lightfoot ntpd[6903]: Listening on interface #2 eth0, 192.168.1.37#123 Enabled
Mar 10 17:25:41 lightfoot ntpd[6903]: synchronized to 194.153.168.75, stratum 2
Mar 10 17:25:41 lightfoot ntpd[6903]: time reset -0.476615 s
Mar 10 17:25:41 lightfoot ntpd[6903]: kernel time sync status change 0001 |
2.6.19-gentoo-r5 | Code: | Mar 10 17:28:29 lightfoot ntpd[7148]: ntpd 4.2.4@1.1437-o Sat Mar 10 16:00:06 UTC 2007 (1)
Mar 10 17:28:29 lightfoot ntpd[7153]: precision = 1.000 usec
Mar 10 17:28:29 lightfoot ntpd[7153]: Listening on interface #0 wildcard, 0.0.0.0#123 Disabled
Mar 10 17:28:29 lightfoot ntpd[7153]: Listening on interface #1 lo, 127.0.0.1#123 Enabled
Mar 10 17:28:29 lightfoot ntpd[7153]: Listening on interface #2 eth0, 192.168.1.37#123 Enabled
Mar 10 17:28:29 lightfoot ntpd[7153]: kernel time sync status 0040
Mar 10 17:28:29 lightfoot ntpd[7153]: frequency initialized 57.084 PPM from /var/lib/ntp/ntp.drift
Mar 10 17:32:48 lightfoot ntpd[7153]: synchronized to 195.177.253.180, stratum 3
Mar 10 17:32:47 lightfoot ntpd[7153]: time reset -0.503255 s
Mar 10 17:32:47 lightfoot ntpd[7153]: kernel time sync status change 0001 |
and 2.6.20-gentoo-r2 | Code: | Mar 10 17:39:17 lightfoot ntpd[6814]: ntpd 4.2.4@1.1437-o Sat Mar 10 16:00:06 UTC 2007 (1)
Mar 10 17:39:17 lightfoot ntpd[6815]: precision = 1.000 usec
Mar 10 17:39:17 lightfoot ntpd[6815]: Listening on interface #0 wildcard, 0.0.0.0#123 Disabled
Mar 10 17:39:17 lightfoot ntpd[6815]: Listening on interface #1 lo, 127.0.0.1#123 Enabled
Mar 10 17:39:17 lightfoot ntpd[6815]: kernel time sync status 0040
Mar 10 17:39:17 lightfoot ntpd[6815]: frequency initialized 57.088 PPM from /var/lib/ntp/ntp.drift
Mar 10 17:39:18 lightfoot ntpd[6815]: Listening on interface #2 eth0, 192.168.1.37#123 Enabled
Mar 10 17:44:17 lightfoot ntpd[6815]: synchronized to 193.201.200.74, stratum 4
Mar 10 17:44:16 lightfoot ntpd[6815]: time reset -0.501546 s
Mar 10 17:44:16 lightfoot ntpd[6815]: kernel time sync status change 0001 |
_________________
The avatar is jorma, a "duck" from "Elephants Dream": the film and all the production materials have been made available under a Creative Commons Attribution 2.5 License, see orange.blender.org for details. |
|
| Back to top |
|
|
cmmalone
n00b
Joined: 08 Oct 2004
Posts: 12
|
| Posted: Sat Mar 10, 2007 10:07 pm Post subject: |
|
|
Add
to your package.use, maybe? |
|
| Back to top |
|
|
AaronPPC
Guru
Joined: 29 May 2005
Posts: 522
Location: Tucson, AZ
|
| Posted: Sun Mar 11, 2007 12:28 am Post subject: |
|
|
Yup. It's fixed here, thanks. I've been watching this thread with great interest.
_________________
--Aaron |
|
| Back to top |
|
|
blaize
n00b
Joined: 15 Mar 2006
Posts: 6
Location: Moscow, Russian Federation
|
| Posted: Sun Mar 11, 2007 12:14 pm Post subject: |
|
|
My ntp exited in daemon mode, but with options "ntp -n" worked fine.
I emerged it with USE="caps" and it is ok with daemon mode now.
Thanks. 
PS: uname -a: Linux blaize 2.6.15-gentoo-r7-gds #1 SMP Sun Apr 9 14:01:52 MSD 2006 x86_64 Intel(R) Xeon(TM) CPU 2.80GHz GenuineIntel GNU/Linux |
|
| Back to top |
|
|
snIP3r
l33t
Joined: 21 May 2004
Posts: 853
Location: germany
|
| Posted: Sun Mar 11, 2007 12:45 pm Post subject: |
|
|
hi!
i got the same error. but after switching back to 4.2.2p3 everything runs as normal!
hope the next version of ntp will fix this error...
HTH
snIP3r
_________________
Intel i3-4130T on ASUS P9D-X
Kernel 5.15.88-gentoo SMP
-----------------------------------------------
if your problem is fixed please add something like [solved] to the topic! |
|
| Back to top |
|
|
AllanS
n00b
Joined: 19 Oct 2006
Posts: 48
|
| Posted: Sun Mar 11, 2007 2:57 pm Post subject: |
|
|
Adding "debug" to USE did not add any information to the problem - ntpd still failed to start.
However, adding "caps" to my USE did fix my problem and now ntpd starts up and works. A subtle solution to a strange problem - nicely caught!
Thanks! |
|
| Back to top |
|
|
deoren
Apprentice
Joined: 23 Dec 2004
Posts: 227
Location: USA
|
| Posted: Wed Jul 11, 2007 1:59 pm Post subject: |
|
|
Anyone still having problems?
I tried the recent stable release of net-misc/ntp-4.2.4_p0 and the problem resurfaced for me.
With the recent removal of net-misc/ntp-4.2.2_p3 from the portage tree (and my 'inability' to actually make sure I had a downgrade path) it looks like an upgrade from the 2.4 kernel series may be in order.
Upstream has marked this as a kernel error and the Gentoo bug entry for this shows it resolved upstream.
Anyone know what that final resolution is?
Thanks. |
|
| Back to top |
|
|
|
Networking & Security
