RTC clock on instance is 1 second faster then Local time
Linode
Linode Staff
The RTC clock on the instance is 1 second faster then the local time / UTC that is causing our AWS Congito function to fail.
Would it be possible to sync the RTC clock or are there other work arounds?
2 Replies
rgerke
Linode Staff
I did a bit of research for you on this and came up with a couple of resources that should be able to help you out. I'll link them below.
The first deals with NTP, and it goes over how to synchronize your system clock automatically:
How to Synchronize Time with NTP in Linux
This next page shows you how to do this with both NTP and RTC:
Set Time Zone and Synchronize System Clock to Your Time Zone in Linux
Some very old posts seem to imply that the clock on guest instances is continually or periodically synced from the host, not just at boot but also while the guest is running. But I have not been able to find a definitive answer. Can anyone confirm or deny this ?
I run the ntp daemon on my instances, with a fixed set of ntp servers, ie. not the dynamically resolved pooled servers. I see some strange behavior where the offset from all of the servers slowly increases from almost zero to tens of milliseconds, then once a week (I think) drops back to ~ 0. As far as I know ntp left to itself should not behave this way, so I do suspect some interference from the host.
--
Ian