Bug 2080056

Summary:	[RFE] move the default timeservers for chrony into /etc/chrony/vendor-timeservers.conf
Product:	[Fedora] Fedora	Reporter:	Pat Riehecky <riehecky>
Component:	chrony	Assignee:	Miroslav Lichvar <mlichvar>
Status:	NEW ---	QA Contact:	Fedora Extras Quality Assurance <extras-qa>
Severity:	unspecified	Docs Contact:
Priority:	unspecified
Version:	rawhide	CC:	mlichvar
Target Milestone:	---
Target Release:	---
Hardware:	Unspecified
OS:	Unspecified
Whiteboard:
Fixed In Version:		Doc Type:	If docs needed, set a value
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:		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 Pat Riehecky 2022-04-28 20:59:14 UTC

Description of problem:

For downstream customization, it is nice to inherit the default config of settings.  However, for my site I'd like to set the system to use my local timeservers and not to bother the default pool.

When I edit /etc/chrony.conf and remove pool, rpm correctly detects I changed a config and any new configs are stored as a `.rpmbak` file.

If the timeservers were in another file, I could more easily separate my change to the pool from the other defaults.


Version-Release number of selected component (if applicable):chrony-4.2-1.fc35.x86_64


How reproducible:
100%


Steps to Reproduce:
1. Note the pool is set in /etc/chrony.conf
2.
3.

Actual results:
Note the pool is set in /etc/chrony.conf

Expected results:
The pool is in some sort of separate file in /etc/chrony.d/


Additional info:

Comment 1 Miroslav Lichvar 2022-05-04 12:49:29 UTC

There were some requests to split the configuration before. See bug #1956320 for an example.

I also asked on the Fedora devel list:
https://www.mail-archive.com/devel@lists.fedoraproject.org/msg155794.html

The trouble is that it would break existing tools like anaconda or augeas that can modify the chrony configuration. If the servers were specified in a separate file, they would not see them and this could lead to unexpected configurations. The tools would need to know how the files are included and what needs to be modified/overriden. That's a major complication.

As the default config is very simple (it has only 3 effective directives beside the pool), I'm not sure if isolating the sources is worth the trouble. I'd say it is easier to create the config from scratch (e.g. using an ansible role).