Note: This bug is displayed in read-only format because the product is no longer active in Red Hat Bugzilla.

Bug 1194571

Summary: instack does not set up ntp on nodes
Product: Red Hat OpenStack Reporter: Dan Sneddon <dsneddon>
Component: instackAssignee: James Slagle <jslagle>
Status: CLOSED DUPLICATE QA Contact: Marius Cornea <mcornea>
Severity: unspecified Docs Contact:
Priority: unspecified    
Version: 6.0 (Juno)CC: calfonso, gfidente, jslagle, mburns, rhel-osp-director-maint
Target Milestone: gaKeywords: TestOnly, Triaged
Target Release: DirectorFlags: yeylon: needinfo+
Hardware: Unspecified   
OS: Unspecified   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2015-06-29 11:49:15 UTC Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:

Description Dan Sneddon 2015-02-20 09:05:46 UTC
Description of problem:
When deploying RHOS6 with instack, the nodes are not set up with ntpd, so the system times are wildly off.

Version-Release number of selected component (if applicable):
1.0.42-1.el7ost

How reproducible:
100%

Steps to Reproduce:
1. Install instack-undercloud
2. instack deploy-overcloud
3. 

Actual results:
None of the deployed nodes have ntpd configured

Expected results:
The nodes should have ntpd configured, and the ntp servers should be a parameter

Additional info:

Comment 3 Giulio Fidente 2015-02-20 16:22:07 UTC
I am seeing this problem as well, when NtpServer is empty the ntp.conf file is misconfigured with a line starting with:

server [blank]

with no hostname or ip appended. This was fixed upstream by a few changes:

https://github.com/openstack/tripleo-image-elements/commit/b8266fa7ecab87a4ae015139e8f8b8a7cd5c714e

https://github.com/openstack/tripleo-image-elements/commit/b8266fa7ecab87a4ae015139e8f8b8a7cd5c714e

https://github.com/openstack/tripleo-heat-templates/commit/42a3b362479f0f9e55f8f57f7fd6d638e5bf36bf

Comment 5 chris alfonso 2015-06-15 19:26:37 UTC
Probably needs docs on how to configure it (adding the right params)

Comment 7 Dan Sneddon 2015-06-18 20:45:48 UTC
I am getting reports that HA does not function if NTP is not synced. Galera will refuse to achieve quorum if the time is too far off. We really need to document not only the configuration of NTP, but also how to make sure that the server has a route to the NTP server.

We might want to roll and NTP server into the undercloud to use as a default or for POC deployments. I'll open a separate BZ for that RFE.

Comment 8 Marius Cornea 2015-06-18 20:59:14 UTC
Ceph cluster also shows some warning about clock skew:

[heat-admin@overcloud-controller-0 ~]$ sudo ceph status
    cluster 17e90c5a-15d3-11e5-ba4a-5254006620d7
     health HEALTH_WARN
            clock skew detected on mon.overcloud-controller-2, mon.overcloud-controller-0
 
[heat-admin@overcloud-controller-0 ~]$ sudo ceph health detail
HEALTH_WARN clock skew detected on mon.overcloud-controller-2, mon.overcloud-controller-0
mon.overcloud-controller-2 addr 192.0.2.20:6789/0 clock skew 0.60983s > max 0.05s (latency 0.00243356s)
mon.overcloud-controller-0 addr 192.0.2.24:6789/0 clock skew 0.236338s > max 0.05s (latency 0.0137471s)

Comment 9 Marius Cornea 2015-06-29 09:16:12 UTC
Where can I set the ntp server(s) so I can veryfiy this? Thanks

Comment 10 James Slagle 2015-06-29 11:49:15 UTC
(In reply to Marius Cornea from comment #9)
> Where can I set the ntp server(s) so I can veryfiy this? Thanks

this will be fixed as part of https://bugzilla.redhat.com/show_bug.cgi?id=1233916
once that lands, ntp server can be set via the --ntp-server argument to openstack overcloud deploy

*** This bug has been marked as a duplicate of bug 1233916 ***