Incorrect timestamp in log entries (seconds: -1)
We sometimes see log entries like the following:
{
"severity": "info",
"source": "kernel",
"content": "Booting Linux on physical CPU 0x0000000000 [0x410fd083]\n",
"msgid": 1,
"timestamp": {
"seconds": -1,
"nanos": 500806665,
"str": "1969-12-31 11:59:59.500806665"
}
}
The timestamp appears to be incorrect; seconds: -1 seems suspicious. As far as I understand, kernel messages are handled by newlogd and the timestamp to JSON is added there. This issue should be rechecked to ensure accurate logging.
Recent release
14.1.0
@deitch any ideas? =)
@deitch any ideas? =)
@deitch seconding this request, on behalf of @OhmSpectator. I will be working on this issue.
@OhmSpectator is this reproducible?
@OhmSpectator is this reproducible?
I remember that I saw it pretty in all the logs at the moment I created the issue. However, I did not attempt to reproduce it intentionally.