Clock drift across your hosted CircleCI installation's instances can cause a wide range of issues, from failed health checks, to failed jobs. You might also notice AWS credential errors like below.
2020-08-05T15:42:43.461439671Z Caused by: com.amazonaws.services.ec2.model.AmazonEC2Exception: AWS was not able to validate the provided access credentials (Service: AmazonEC2; Status Code: 401; Error Code: AuthFailure; Request ID
Typically, error messages will indicate something about certificates "not yet being valid," and jobs will display curious runtimes ("finished five minutes from now").
Here's how to re-sync clocks across your instances:
- SSH into the Services box
- Install ntp if you haven't already:
sudo apt-get install ntp
- Restart ntp. These steps may need to be run as root:
service ntp stop
ntpdate -s time.nist.gov
service ntp start