Bug 2063953

Summary:	extend rest mappings with 'job' definition
Product:	OpenShift Container Platform	Reporter:	Sean Merrow <smerrow>
Component:	kube-apiserver	Assignee:	Abu Kashem <akashem>
Status:	CLOSED ERRATA	QA Contact:	Mike Fiedler <mifiedle>
Severity:	high	Docs Contact:
Priority:	high
Version:	4.8	CC:	akashem, aos-bugs, chris.brown, jmekkatt, josephw, kewang, mfojtik, michal, mifiedle, nsarath, sleigon, smozowei, sttts, vlaad, wking, wlewis, xxia
Target Milestone:	---	Keywords:	FastFix
Target Release:	4.8.z
Hardware:	All
OS:	Linux
Whiteboard:	EmergencyRequest
Fixed In Version:		Doc Type:	If docs needed, set a value
Doc Text:		Story Points:	---
Clone Of:	2048801
Clones:	2069310 (view as bug list)		Environment:
Last Closed:	2022-04-27 11:46:15 UTC	Type:	---
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	---	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:
Embargoed:
Bug Depends On:	2069310
Bug Blocks:

Description Sean Merrow 2022-03-14 17:29:18 UTC

+++ This bug was initially created as a clone of Bug #2048801 +++

Description of problem:

Need to implement https://github.com/openshift/kubernetes/pull/1209 in order to extend rest mappings with job definition. This is because when the kube-api-server loses connecitivity to the pod networking, a 'job' cannot be run. The current list of resources in the RESTmappings includes many resources, but not 'jobs'


How reproducible:
Can reproduce every time.


Steps to Reproduce:
1. On one of the master nodes, remove route to pod network, then try to run a 'job'


Actual results:
job fails


Expected results:
job succeeds

Comment 4 Abu Kashem 2022-04-15 00:09:26 UTC

it's waiting on qe to verify 4.9, once 4.9 is verified by qe, we will open a PR on openshift/kubernetes for 4.8. (the delay is due to the fact that we have to make this fix from master branch, and all the down to 4.8, one at a time).

4.19 PR: https://github.com/openshift/kubernetes/pull/1233.

Comment 5 Mike Fiedler 2022-04-18 17:52:09 UTC

Verified pre-merge using a cluster-bot cluster built from  https://github.com/openshift/kubernetes/pull/1245  using steps here:  https://bugzilla.redhat.com/show_bug.cgi?id=2069310#c5.   Job completed successfully and oc get jobs show successful completion.

[root@ip-172-31-18-207 work]# oc get jobs
NAME   COMPLETIONS   DURATION   AGE
mypi   1/1           33s        64s
[root@ip-172-31-18-207 work]# oc get pods
NAME         READY   STATUS      RESTARTS   AGE
mypi-npr79   0/1     Completed   0          96s

Marking PR qe-approved and this bz as tested.

Comment 7 W. Trevor King 2022-04-20 14:12:17 UTC

I'm dropping the master pull [1] from this 4.8.z bug.

[1]: https://github.com/openshift/kubernetes/pull/1209

Comment 11 Mike Fiedler 2022-04-20 15:37:29 UTC

Verified on 4.8.38 candidate using verification procedure at https://bugzilla.redhat.com/show_bug.cgi?id=2069310#c5

job runs successfully after deleting pod network route on one master
oc get jobs correctly shows job status

Comment 13 errata-xmlrpc 2022-04-27 11:46:15 UTC

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 (OpenShift Container Platform 4.8.39 bug fix update), 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://access.redhat.com/errata/RHBA-2022:1427