clock inside rancher docker vm is running

[biotecs]

Hey everyone,

First of all my FreeNAS currently looks like this:

OS version: FreeNAS-11.2-U5
(Build Date: Jun 24, 2019 18:41)
Processor: Intel(R) Pentium(R) CPU N3700 @ 1.60GHz (4 cores)
Memory: 15 GiB

For some days now I have been dealing with a time problem within my rancher docker vm. At first I thought of a time zone problem, but in the meantime I found out that it is a running clock. After the vm has started and the ntp system-docker has picked up the time, the clock starts to drift. Within 10 minutes that's already several seconds, so that on the day it's even minutes to hours. The internal ntp does not seem to arrive against it, while the time of the FreeNAS host itself remains clean and stable.

I've already done some research on the net, both here in the forum, where I found this post:

https://www.ixsystems.com/community/threads/emulated-clock-in-bhyve-is-local-time.59266/

as well as in the Rancher Forum, where you can read this among others:

https://forums.rancher.com/t/keeping-the-time-correct-on-rancheros/10520

Also in the issues of RancherOS there were topics in this area from time to time, but they are all considered fixed, like for example:

https://github.com/rancher/os/issues/1963

Last but not least, I have to admit that I can't get any further now. I have already tried to see if there is a problem with the system-docker, but it seems to work normally except for a Clock Unsynchronized Error. A system-docker logs ntp shows:

Show : system-docker logs ntp

2019/08/2011 20:46:27 Writing file to "/etc/ntp.conf"
2019/08/11 20:46:27rote file to "/etc/ntp.conf"
time="2019-08-11T20:46:27Z" level=info msg="Wrote file /etc/ntp.conf to filesystem"
+ echo 'starting in one shot mode to fix large time differences'
starting in one shot mode to fix large time differences
+ ntpd -gq
11 Aug 20:46:27 ntpd[14]: ntpd 4.2.8p12@1.3728-o Sat Feb 23 20:00:00 UTC 2019 (1): Starting
11 Aug 20:46:27 ntpd[14]: Command line: ntpd -gq
11 Aug 20:46:27 ntpd[14]: bind(20) AF_INET6 fe80::2a0:98ff:fe42:d921%2#123 flags 0x11 failed: Cannot assign requested address
11 Aug 20:46:27 ntpd[14]: unable to create socket on eth0 (4) for fe80::2a0:98ff:fe42:d921%2#123
11 Aug 20:46:27 ntpd[14]: proto: precision = 0.270 usec (-22)
11 Aug 20:46:27 ntpd[14]: Listen and drop on 0 v6wildcard [::]:123
11 Aug 20:46:27 ntpd[14]: Listen and drop on 1 v4wildcard 0.0.0.0:123
11 Aug 20:46:27 ntpd[14]: Listen normally on 2 lo 127.0.0.1:123
11 Aug 20:46:27 ntpd[14]: Listen normally on 3 lo [::1]:123
11 Aug 20:46:27 ntpd[14]: failed to init interface for address fe80::2a0:98ff:fe42:d921%2
11 Aug 20:46:27 ntpd[14]: Listening on routing socket on fd #20 for interface updates
11 Aug 20:46:30 ntpd[14]: Listen normally on 5 eth0 192.168.2.60:123
11 Aug 20:46:30 ntpd[14]: Listen normally on 6 eth0 [fe80::2a0:98ff:fe42:d921%2]:123
11 Aug 18:46:38 ntpd[14]: ntpd: time set -7205.234024 s
ntpd: time set -7205.234024s
+ echo 'starting long running nptd'
starting long running nptd
+ exec ntpd --nofork -g

Maybe someone has a similar problem and was able to solve it somehow, I would be grateful for any help.

 

[dlavigne]

Were you able to resolve this?

 

[biotecs]

Hey Dru,

unfortunatly I wasn‘t able to fix it. But I found out that if I deploy another os (e.g. Archlinux) within a standard FreeNAS virtual machine the clock is find. So it seems that the problem only exists with the rancher Docker Host.