19 Oct
2020
19 Oct
'20
8:35 p.m.
On 19/10/2020 21:35, Paul Lewis wrote:
Hi all,
I’ve noticed that over the past few reboot cycles for security patches,
my VM suspends and restores fine, and all services restore fine except
ntp, which never recovers. When checking the status of the service I get:
I've recently experienced similar behaviour when ntp "lost it" if the
upstream servers disappeared and came back.
Part of the problem seems to be that ntp only resolves servers once and
if (for absolutely innocent and sensible reasons) ntp is started before
there's dns and net available, it'll break.
However, even when we "fixed" this by using ip-addresses instead we'd
see ntp losing sync and die for apparently no reason.
We solved it by changing to chrony which both syncs up faster and
apparently keeps better time.
- OM
root@jaguar:~# service ntp status
● ntp.service - Network Time Service
Loaded: loaded (/lib/systemd/system/ntp.service; enabled; vendor
preset: enabled)
Active: inactive (dead)
Docs: man:ntpd(8)
Restarting the service restores everything back to working order. But I
can’t always guarantee that I can get to the console quickly, so the
icinga email alerts keep rolling in…
If it makes a difference, I’m running Ubuntu 18.04.4, and
ntp 1:4.2.8p10+dfsg-5ubuntu7.3.
I presume the issue is the huge jump in time the kernel/ntp service sees
when the VM is restored, is there a good way of getting ntpd to handle
this? Do other people see this issue, and if so, what
solutions/workarounds do you use to prevent it happening?
Thanks,
Paul
_______________________________________________
users mailing list
users(a)lists.bitfolk.com
https://lists.bitfolk.com/mailman/listinfo/users