clock drifting like mad

Hello,

I have tumbleweed on 2 unrelated PCs, 1 laptop and 1 desktop, none of them goes into hibernation and none of them has dual boot. Since something like 2 weeks, the clock drifts a lot on both, like 1 min per 10 min. Using ntpd is not sufficient to keep the time, I had to set up a cronjob every minute running ntpdate to be able to keep a more or less consistent time.
Those PCs are both using different ntp servers, are not used at the same location, have totally different hardware specs and are from completely different manufacturers. The only common point is that they are both running tumbleweed.

Any idea?

Thanks

I installed openntpd and it seems that one is able to keep the clock.

Ah, non, even with openntpd it doesn’t work. It seems it’s able to set the clock once, but not more.
both the original ntp and ntpdat tell me they are unable to get a valid timeserver (and it’s a valid stratum 1 I am trying to sync with)

2015-01-30T08:14:07.650010+01:00 bilbo ntpd[31506]: set local clock to Fri Jan 30 08:14:07 CET 2015 (offset -4753.067949s)
2015-01-30T08:14:28.674180+01:00 bilbo ntpd[31507]: peer 172.20.0.254 now valid
2015-01-30T08:15:21.730036+01:00 bilbo ntpd[31509]: adjusting local clock by -2.590253s
2015-01-30T08:18:29.897449+01:00 bilbo ntpd[31509]: adjusting local clock by -17.859443s
2015-01-30T08:22:52.139824+01:00 bilbo ntpd[31509]: adjusting local clock by -27.114457s
2015-01-30T08:24:57.265078+01:00 bilbo ntpd[31509]: adjusting local clock by -31.387239s
2015-01-30T08:26:02.305321+01:00 bilbo ntpd[31509]: adjusting local clock by -37.416031s
2015-01-30T08:27:42.367246+01:00 bilbo ntpd[31509]: adjusting local clock by -54.665301s
2015-01-30T08:31:58.578857+01:00 bilbo ntpd[31509]: adjusting local clock by -73.163075s
2015-01-30T08:36:09.817324+01:00 bilbo ntpd[31509]: adjusting local clock by -79.623373s
2015-01-30T08:37:47.917750+01:00 bilbo ntpd[31509]: adjusting local clock by -86.424523s
2015-01-30T08:39:24.987352+01:00 bilbo ntpd[31509]: adjusting local clock by -102.018935s
2015-01-30T08:43:05.162884+01:00 bilbo ntpd[31509]: adjusting local clock by -112.582783s
2015-01-30T08:45:46.292942+01:00 bilbo ntpd[31509]: adjusting local clock by -126.265452s
2015-01-30T08:48:56.477518+01:00 bilbo ntpd[31509]: adjusting local clock by -135.640488s
2015-01-30T08:51:03.535624+01:00 bilbo ntpd[31509]: adjusting local clock by -152.370313s

I can only think of two possible explanations:

(1) a serious hardware problem;
(2) a hidden network proxy is handling your ntp requests and feeding bogus data.

If your system keeps good time without “ntpd”, then the problem is the second of those. If it cannot keep reasonably good time without “ntpd” then you have a hardware problem. But I would expect a hardware problem to cause other symptoms.

The second proposal is impossible as the desktop is getting its time from my gateway which uses a DCF77 clock receiver to sync time.

On the desktop however, it might indeed be a hardware problem as that machine is quite old.

The strange thing is that I got the same issue on 2 different unrelated machines running on different networks and syncing with different local ntp servers.