kernel time is difference from system time

jogyulas · 06-14-2018, 01:07 AM

Hi Guys,

I have a question. I have just noticed in logging files that my kernel time is difference from system time. Is it okay or not? I tried to solve this problem by google but I didnt find solution for it.

Can you help me and/or explain me why is not same?

Ty,

mrmazda · 06-14-2018, 01:41 AM

Is your PC multiboot with Windows? What does /etc/adjtime contain? LOCAL rather than UTC should explain the difference.

jogyulas · 06-14-2018, 02:12 AM

Hi mrmazda,

Ty for your quick response. The Debian OS runs on a server so it is a single boot.

Code:

cat /etc/adjtime
0.000010 1528955839 0.000000
1528955839
UTC

date
Thu Jun 14 09:08:12 CEST 2018

Q2: do you have a Mazda?

mrmazda · 06-14-2018, 02:39 AM

What does /etc/localtime link to? What is your local time zone, +0200? Is that how much difference you see in the logs?

X-LFS-2010 · 06-14-2018, 03:02 AM

it's not called kernel time and system time for no reason: they will differ a _little

the system (libc) gets time from the kernel at the time the application asks for the time, the kernel takes time from the hardware (most kernels are configured that way, anyhow)

it's possible for an application or even libc to get delayed in showing you the time after it obtains it from the kernel

jogyulas · 06-14-2018, 04:12 AM

@mrmazda: if I understand well what you would like to see here it is:

Code:

ls -l /etc/localtime
-rw-r--r-- 1 root root 2379 Jun 14 07:47 /etc/localtime

I have also tried to reconfigure my timezone with tzdata, but my system time is ok

Code:

Current default time zone: 'Europe/Budapest'
Local time is now:      Thu Jun 14 10:57:32 CEST 2018.
Universal Time is now:  Thu Jun 14 08:57:32 UTC 2018.

Difference between the two times is about 30 minutes but it isnt exact. And just to present the problem it is a snippet from my syslog file:

Code:

Jun 14 11:00:01 x zfs-snapshot: executing zfs snapshot pool/storage@auto-201806140900-UTC
Jun 14 11:00:02 x zfs-snapshot: executing zfs destroy pool/storage@auto-201806070900-UTC
Jun 14 10:37:55 x kernel: tg3 0000:03:00.0 eth0: Link is down
Jun 14 10:37:59 x kernel: tg3 0000:03:00.0 eth0: Link is up at 1000 Mbps, full duplex
Jun 14 10:37:59 x kernel: tg3 0000:03:00.0 eth0: Flow control is on for TX and on for RX
Jun 14 10:37:59 x kernel: tg3 0000:03:00.0 eth0: EEE is enabled
Jun 14 11:06:03 x syslog-ng[52107]: syslog-ng shutting down; version='3.5.6'
Jun 14 11:06:03 x syslog-ng[3008]: syslog-ng starting up; version='3.5.6'

As you can see logging from kernel uses different timestamps. It happens too when iptables logs into other logging files.

@X-LFS-2010: Ty for your response too
I understand what you write and I have checked my server time on iLO which shows the right time (aside from +2 hours):

Code:

iLO Date/Time:	Thu Jun 14 08:57:32 2018

jogyulas · 06-19-2018, 08:29 AM

Hi,

It seems I have found it, and it also seems that this problem is caused by syslog-ng. Perhaps it is a bug :/ I need to upgrade syslog-ng pkg.