Bug 1880318

Summary: [release 4.5] insights-operator: Fix bug in reflector not recovering from "Too large resource version"
Product: OpenShift Container Platform Reporter: Lukasz Szaszkiewicz <lszaszki>
Component: Insights OperatorAssignee: Tomas Remes <tremes>
Status: CLOSED ERRATA QA Contact: Pavel Šimovec <psimovec>
Severity: high Docs Contact:
Priority: high    
Version: 4.5CC: aos-bugs, inecas, palonsor, tremes
Target Milestone: ---Keywords: Reopened
Target Release: 4.5.z   
Hardware: Unspecified   
OS: Unspecified   
Whiteboard:
Fixed In Version: Doc Type: No Doc Update
Doc Text:
Story Points: ---
Clone Of:
: 1882073 (view as bug list) Environment:
Last Closed: 2021-01-20 05:49:27 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:
Bug Depends On: 1880369    
Bug Blocks: 1879901    

Description Lukasz Szaszkiewicz 2020-09-18 09:47:48 UTC
A recent fix in the reflector/informer https://github.com/kubernetes/kubernetes/pull/92688 prevents components/operators from entering a hotloop and stuck.

There are already reported cases that have run into that issue and were stuck for hours or even days. For example https://bugzilla.redhat.com/show_bug.cgi?id=1877346.

The root cause of the issue is the fact that a watch cache is initialized from the global revision (etcd) and might stay on it for an undefined period (if no changes were (add, modify) made). 
That means that the watch cache across server instances may be out of sync. 
That might lead to a situation in which a client gets a resource version from a server that has observed a newer rv, disconnect (due to a network error) from it, and reconnect to a server that is behind, resulting in “Too large resource version“ errors.

More details in https://github.com/kubernetes/kubernetes/issues/91073 and https://github.com/kubernetes/enhancements/tree/master/keps/sig-api-machinery/1904-efficient-watch-resumption


It looks like the issue only affects 1.18. According to https://github.com/kubernetes/kubernetes/issues/91073#issuecomment-652251669 the issue was first introduced in that version by changes done to the reflector. 
The fix is already present in 1.19.


Please make sure that insights-operator is using a client-go that includes https://github.com/kubernetes/kubernetes/pull/92688 if not please use this BZ and file a PR.
In case you are using a framework to build your operator make sure it uses the right version of the client-go library.

Comment 2 Martin Kunc 2020-09-24 04:31:24 UTC
Insights Operator in 4.5 doesn't use affected client-go. We are still on v.0.17.1

Comment 3 Lukasz Szaszkiewicz 2020-09-24 07:32:14 UTC
(In reply to Martin Kunc from comment #2)
> Insights Operator in 4.5 doesn't use affected client-go. We are still on
> v.0.17.1

OpenShift in 4.5 can send a msg that the informers must understand and react.
We need to make sure that operators in 4.5 are at least on 1.18.6.

Comment 5 Lukasz Szaszkiewicz 2020-12-04 11:50:26 UTC
any update? what's the current status?

Comment 6 Tomas Remes 2020-12-04 12:43:49 UTC
I will check this one next week. The desired version for k8s api and client-go is still v0.18.6 right?

Comment 7 Lukasz Szaszkiewicz 2020-12-04 12:47:29 UTC
(In reply to Tomas Remes from comment #6)
> I will check this one next week. The desired version for k8s api and
> client-go is still v0.18.6 right?

thanks, correct that the minimum version that has the fix.

Comment 10 Pavel Šimovec 2021-01-13 14:45:17 UTC
k8s api upgraded to the required version - k8s.io/api v0.18.6

Comment 12 errata-xmlrpc 2021-01-20 05:49:27 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 4.5.27 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-2021:0033