Description of problem: ntp server is not configured on controller nodes even though --ntp-server was specified but it is configured on computes. This has for effect that controller time might skew and generate issues Version-Release number of selected component (if applicable): How reproducible: RHOSP 7.0 and RHOSP 8.0 Steps to Reproduce: 1. Deploy the overcloud with --ntp-server 2. Watch for /etc/ntp.conf 3. Actual results: Not configured on controllers Expected results: Should be configured everywhere and at this point, why not configure ntp.conf on controller to syn with external source and sync all computes with the controllers as source? Additional info:
This bug did not make the OSP 8.0 release. It is being deferred to OSP 10.
Same issue on current OSP10 last rhos release, on all nodes not only the controllers. Tried successfully on OSP9 though.
Used last rhos-release puddle, seems fixed now. Deployed with --ntp-server 10.16.255.1 [root@overcloud-controller-0 ~]# grep server /etc/ntp.conf server 10.16.255.1 [root@overcloud-controller-0 ~]# ntpq -p remote refid st t when poll reach delay offset jitter ============================================================================== 10.16.255.1 .XFAC. 16 u 68 64 0 0.000 0.000 0.000 Cheers, Greg
On OSP10: [root@overcloud-controller-0 heat-admin]# grep ^server /etc/ntp.conf server clock.redhat.com server clock.redhat.com which matches the NTP servers set in an environment file: [stack@undercloud-0 ~]$ grep NtpServer network-environment.yaml NtpServer: ["clock.redhat.com","clock.redhat.com"]
Verified with --ntp-server as well. Any specific reason for not using chronyd ?
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://rhn.redhat.com/errata/RHEA-2016-2948.html