Note: This bug is displayed in read-only format because the product is no longer active in Red Hat Bugzilla.

Bug 1792533

Summary:	OCP 4.2.12: ingress and network operators degraded after upgrade to 4.3 (leaking on kubelet restart)
Product:	OpenShift Container Platform	Reporter:	Ben Bennett <bbennett>
Component:	Networking	Assignee:	Alexander Constantinescu <aconstan>
Networking sub component:	openshift-sdn	QA Contact:	zhaozhanqi <zzhao>
Status:	CLOSED NOTABUG	Docs Contact:
Severity:	high
Priority:	unspecified	CC:	aconstan, aos-bugs, bbennett, ccoleman, cdc, danw, dosmith, huirwang, jokerman, mifiedle, mpatel, rphillips, scuppett, wabouham, weliang, wking, wsun, xtian, zzhao
Version:	4.3.0
Target Milestone:	---
Target Release:	4.4.0
Hardware:	x86_64
OS:	Linux
Whiteboard:
Fixed In Version:		Doc Type:	If docs needed, set a value
Doc Text:		Story Points:	---
Clone Of:	1787635
Clones:	1792536 (view as bug list)		Environment:
Last Closed:	2020-02-21 10:14:43 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:
Bug Blocks:	1792536

Comment 1 Ben Bennett 2020-01-17 21:08:12 UTC

I've cloned 1787635 to capture that we leak under two circumstances:
1) When a node was rebooted
2) When a kubelet was restarted 

This bug will cover case 2, the kubelet restart.

Bug 1787635 is only for the node reboot leak.

Comment 2 Weibin Liang 2020-02-18 00:02:53 UTC

@Alexander

There no 4.3.0-0.nightly-2020-01-02-141332 any more in nightly build, can not reproduce from OCP 4.2.12 to 4.3.0-0.nightly-2020-01-02-141332

Tried upgrading from OCP 4.2.12 to 4.3.2 twice in two clusters, upgrading works fine.

Comment 3 Alexander Constantinescu 2020-02-18 10:57:31 UTC

@Ben

Please advise on these bugs: this one and https://bugzilla.redhat.com/show_bug.cgi?id=1792536

This one covers kubelet restarts from what you mentioned in #comment 1, and https://bugzilla.redhat.com/show_bug.cgi?id=1792536 covers node reboot - however: http://bugzilla.redhat.com/show_bug.cgi?id=1792536#c1 indicates that the node reboot scenario is fixed and handled by https://bugzilla.redhat.com/show_bug.cgi?id=1787635. Should we close 1792536 then?

Does this bug even have anything to do with the title and description anymore - i.e upgrades from 4.2 -> 4.3? It seems like Weibin should be testing kubelet restarts and checking the directory containing ip address allocations?

/Alex

Comment 4 Weibin Liang 2020-02-18 19:17:02 UTC

@Alexander @Ben

Execute more testing today, here are my testing results:
1. Install v4.2.12 cluster, and deploy some projects and application pods
2. Check ip address allocations, no leaking found
3. Do upgrading from 4.2.12 to 4.3.2
4. Upgrading succeed, no errors for degraded
5. Check ip address allocations, leaking found
6. Run "systemctl restart kubelet" in all three worker nodes
7. Check ip address allocations, no leaking found