Note: This bug is displayed in read-only format because the product is no longer active in Red Hat Bugzilla.

Bug 1728216

Summary:	Operator doesn't update "command" in deployments/daemonsets
Product:	OpenShift Container Platform	Reporter:	Vadim Rutkovsky <vrutkovs>
Component:	Machine Config Operator	Assignee:	Vadim Rutkovsky <vrutkovs>
Status:	CLOSED WONTFIX	QA Contact:	Micah Abbott <miabbott>
Severity:	unspecified	Docs Contact:
Priority:	unspecified
Version:	4.1.z	CC:	eparis, walters
Target Milestone:	---
Target Release:	4.1.z
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:	2019-07-22 19:15:49 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 Vadim Rutkovsky 2019-07-09 10:47:00 UTC

Description of problem:
MCO won't update deployments for MCC / MCD if "command" is changed there. This has been fixed in 4.2 (https://github.com/openshift/machine-config-operator/pull/850), but still valid for 4.1

Version-Release number of selected component (if applicable):
4.1.4

How reproducible:
Always

Steps to Reproduce:
1. oc project openshift-machine-config-operator
2. oc edit ds machine-config-server
3. Modify "args" and "command" in first container

Actual results:
Operator would revert "args" change but won't touch "command" change

Expected results:
Operator would revert "args" and "command" change

Additional info:

Comment 1 Colin Walters 2019-07-09 17:57:55 UTC

I agree this is a bug we should fix, but...what's the argument for backporting this to 4.1?  Risk seems pretty low, admittedly, but so does the benefit.  If someone edits the daemonsets, we tell them "don't do that"...plus it's a fairly easy fix to do manually.

Did a real world customer hit this?

Was this something you hit when doing development?

Comment 2 Vadim Rutkovsky 2019-07-10 08:42:35 UTC

(In reply to Colin Walters from comment #1)
> If someone edits the daemonsets, we tell them "don't do
> that"...plus it's a fairly easy fix to do manually.

Right, but aren't operators supposed to take care of that?

> Did a real world customer hit this?

No, not that I know of at least

> Was this something you hit when doing development?

No, just picked up an upstream issue - https://github.com/openshift/machine-config-operator/issues/93.

Comment 3 Eric Paris 2019-07-22 19:15:49 UTC

Let's just call this WONTFIX and keep charging ahead on 4.2.  Thank you for thinking it through!