Bug 1880337 - [release 4.6] cluster-monitoring-operator: Fix bug in reflector not recovering from "Too large resource version"
Summary: [release 4.6] cluster-monitoring-operator: Fix bug in reflector not recoverin...
Keywords:
Status: CLOSED ERRATA
Alias: None
Product: OpenShift Container Platform
Classification: Red Hat
Component: Monitoring
Version: 4.5
Hardware: Unspecified
OS: Unspecified
high
high
Target Milestone: ---
: 4.6.0
Assignee: Lili Cosic
QA Contact: Junqi Zhao
URL:
Whiteboard:
Depends On: 1880369
Blocks: 1881043 1892585
TreeView+ depends on / blocked
 
Reported: 2020-09-18 10:10 UTC by Lukasz Szaszkiewicz
Modified: 2020-10-29 09:12 UTC (History)
9 users (show)

Fixed In Version:
Doc Type: No Doc Update
Doc Text:
Clone Of:
: 1881043 1881072 1881079 1881107 1881109 1892585 (view as bug list)
Environment:
Last Closed: 2020-10-27 16:42:20 UTC
Target Upstream Version:
Embargoed:


Attachments (Terms of Use)


Links
System ID Private Priority Status Summary Last Updated
Github openshift cluster-monitoring-operator pull 933 0 None closed Bug 1880337: Bump client-go & co. 2020-12-18 12:46:22 UTC
Red Hat Product Errata RHBA-2020:4196 0 None None None 2020-10-27 16:42:45 UTC

Description Lukasz Szaszkiewicz 2020-09-18 10:10:19 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 cluster-monitoring-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 1 Simon Pasquier 2020-09-21 10:46:22 UTC
cluster-monitoring-operator 4.5 depends on k8s.io/client-go v0.17.1 [1] so it isn't affected by this issue. The same goes for prometheus-operator [2].
That being said, the 4.6 branches use k8s.io/client-go v0.18.3 and v0.18.2 [3][4] and they probably need to be fixed. 

@Lukasz Should we open another BZ?

[1] https://github.com/openshift/cluster-monitoring-operator/blob/0c110b7edadad09182983e48013125a07284116d/go.mod#L37
[2] https://github.com/openshift/prometheus-operator/blob/99b893905d26d85d50d1178be195388e5c000322/go.mod#L42
[3] https://github.com/openshift/cluster-monitoring-operator/blob/922578d7d8a33f39b43b577e74c469b4374e90bd/go.mod#L31
[4] https://github.com/openshift/prometheus-operator/blob/52492b3b48ed1e4f851a78a51817e92404cf2767/go.mod#L36

Comment 2 Lukasz Szaszkiewicz 2020-09-21 12:20:10 UTC
(In reply to Simon Pasquier from comment #1)
> cluster-monitoring-operator 4.5 depends on k8s.io/client-go v0.17.1 [1] so
> it isn't affected by this issue. The same goes for prometheus-operator [2].
> That being said, the 4.6 branches use k8s.io/client-go v0.18.3 and v0.18.2
> [3][4] and they probably need to be fixed. 
> 
> @Lukasz Should we open another BZ?
> 
> [1]
> https://github.com/openshift/cluster-monitoring-operator/blob/
> 0c110b7edadad09182983e48013125a07284116d/go.mod#L37
> [2]
> https://github.com/openshift/prometheus-operator/blob/
> 99b893905d26d85d50d1178be195388e5c000322/go.mod#L42
> [3]
> https://github.com/openshift/cluster-monitoring-operator/blob/
> 922578d7d8a33f39b43b577e74c469b4374e90bd/go.mod#L31
> [4]
> https://github.com/openshift/prometheus-operator/blob/
> 52492b3b48ed1e4f851a78a51817e92404cf2767/go.mod#L36

The Kube API in 4.5 is affected. It can return an error that the operators must understand and recover from. Basically anything that uses an informer.
For 4.5/4.6 you should bump at least to 1.18.6 (which has the fix)

Comment 3 Simon Pasquier 2020-09-21 12:30:42 UTC
Targeting this bug against 4.6.0. I'll create a clone for 4.5.z.

Comment 6 Junqi Zhao 2020-09-25 08:34:52 UTC
tested with 4.6.0-0.nightly-2020-09-24-184015, the fix is in the payload and did not see the "Too large resource version" error

Comment 9 errata-xmlrpc 2020-10-27 16:42:20 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 4.6 GA Images), 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-2020:4196


Note You need to log in before you can comment on or make changes to this bug.