[chrony-users] Re: sync takes half an hour to complete

[ Thread Index | Date Index | More chrony.tuxfamily.org/chrony-users Archives ]


Anyone, any ideas?

On Fri, Feb 14, 2020 at 3:24 PM Andrii Voloshyn <a.voloshyn@xxxxxxxxxxxxxx> wrote:
Hi everyone,

     Using quite an old version (chronyc (chrony) version 2.4 (+READLINE +IPV6 -DEBUG)) on an embedded Linux system.
     Recently had one occurrence where the synchronization took almost half an hour to complete.

Some of the output from chronyc waitsync 3 0.5 :

Feb 10 22:27:30 : try: 1, refid: 0.0.0.0, correction: 0.000000035, skew: 0.000
Feb 10 22:27:30 : try: 2, refid: 0.0.0.0, correction: 0.000000035, skew: 0.000
Feb 10 22:27:30 : try: 3, refid: 0.0.0.0, correction: 0.000000035, skew: 0.000
Feb 10 22:27:52 : try: 1, refid: 0.0.0.0, correction: 0.000000035, skew: 0.000
Feb 10 22:27:52 : try: 2, refid: 0.0.0.0, correction: 0.000000035, skew: 0.000
Feb 10 22:27:52 : try: 3, refid: 0.0.0.0, correction: 0.000000035, skew: 0.000
Feb 10 22:28:14 : try: 1, refid: 0.0.0.0, correction: 0.000000035, skew: 0.000
Feb 10 22:28:14 : try: 2, refid: 0.0..0.0, correction: 0.000000036, skew: 0.000
.....
Feb 10 22:43:20 : try: 1, refid: 0.0.0.0, correction: 0.000000039, skew: 0.000
Feb 10 22:43:20 : try: 2, refid: 0.0.0.0, correction: 0.000000039, skew: 0.000
Feb 10 22:43:20 : try: 3, refid: 0.0.0.0, correction: 0.000000039, skew: 0.000
Feb 10 22:43:43 : try: 1, refid: 0.0.0.0, correction: 0.000000039, skew: 0..000
Feb 10 22:43:43 : try: 2, refid: 0.0.0.0, correction: 0.000000039, skew: 0.000
.....
Feb 10 22:58:04 : try: 1, refid: 0.0.0.0, correction: 0.000000003, skew: 0.000
Feb 10 22:58:04 : try: 2, refid: 0.0.0.0, correction: 0.000000003, skew: 0.000
Feb 10 22:58:04 : try: 3, refid: 0.0.0.0, correction: 0.000000003, skew: 0.000
Feb 10 22:58:23 : chronyd[479]: Selected source 81.94.123.16
Feb 10 22:58:26 : try: 1, refid: 0.0.0.0, correction: 0.000000003, skew: 0.000
Feb 10 22:58:26 : try: 2, refid: 0..0.0.0, correction: 0.000000003, skew: 0.000

I thought that correction should be 0.0 while refid is 0.0.0.0?
It looks like it was because it took a long time to select source?, the following line : Feb 10 22:58:23 : chronyd[479]: Selected source 81.94.123.16

Chrony configuration file:

pool <pool server 1>.pool.ntp.org iburst
pool <pool server 2>.pool.ntp.org iburst
pool <pool server 3>.pool.ntp.org iburst
pool <pool server 4>.pool.ntp.org iburst
pool <pool server 5>.pool.ntp.org iburst
makestep 1.0 3

I am wondering if it could be caused by some bug, that might have been fixed in the later versions of chrony?
What could cause it?
I didn't have a chance to run sources or activity command when the problem happened as the device was not attended.
But in the normal operation on the same device, the activity command returns, the following:

20 sources online
0 sources offline
0 sources doing burst (return to online)
0 sources doing burst (return to offline)
0 sources with unknown address

Is there anything I can do to eliminate this issue in the future?
One other question I would like to ask is this: Is the source only selected after all sources get online/offline?


P.S. the Internet connection is established through cellular modem via ppp protocol.

Thank you for your help,
Andrew


Mail converted by MHonArc 2.6.19+ http://listengine.tuxfamily.org/