Bug 1995953
Summary: | Ingresscontroller change the replicas to scaleup first time will be rolling update for all the ingress pods | ||
---|---|---|---|
Product: | OpenShift Container Platform | Reporter: | Jie Wu <jiewu> |
Component: | Networking | Assignee: | Grant Spence <gspence> |
Networking sub component: | router | QA Contact: | Arvind iyengar <aiyengar> |
Status: | CLOSED ERRATA | Docs Contact: | |
Severity: | low | ||
Priority: | medium | CC: | gspence, hongli, mmasters |
Version: | 4.7 | Keywords: | Reopened |
Target Milestone: | --- | ||
Target Release: | 4.11.0 | ||
Hardware: | Unspecified | ||
OS: | Unspecified | ||
Whiteboard: | |||
Fixed In Version: | Doc Type: | No Doc Update | |
Doc Text: | Story Points: | --- | |
Clone Of: | Environment: | ||
Last Closed: | 2022-08-10 10:36:53 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
Jie Wu
2021-08-20 09:35:55 UTC
This BZ is low severity and does not currently have any PR to merge, so we'll push this out to 4.10. Hi, is there any update for fixing the problem of the router ENV sorting entries? Hi Jie, I picked up this bug last week. I'm trying to replicate this bug and having issues doing so. I've tried 4.6.40, 4.7.13, and 4.10.0 versions on IPI (cluster-bot mostly). Every time, the ENV is always sorted with executing the reproducer as you have specified and I don't get a rolling update on initial scaling from 2 to 3. Could I get you to verify if this bug is reproducible on IPI or only with UPI? Hi, Grant The environment is using offline UPI installation. The HAProxy router is using the endpointPublishingStrategy (Other: HostNetwork) binding on 80/443 ports. Could you please re-try to reproduce in UPI environment? Okay - I was able to reproduce with HostNetwork as the endpointPublishingStrategy on a IPI cluster. So it appears the bug is specifically related to ENV sorting with that configuration. I will update you when I am able resolve the issue in code. So I found the root cause of the bug. Like the description mentions, it has been fixed. It was fixed (whether intentionally or not) in 4.7.21 via https://github.com/openshift/cluster-ingress-operator/commit/32df7d28bc1f0ba63eedbbdd763c7124b34ce242. The problem is with the code not callling the deploymentTemplateHash() function specifically for HostNetwork endpointPublishingStrategy types. All other types were fine. The code was updated in 4.7.21 to always call deploymentTemplateHash() which will always sort the environment variables. As far as I know, we aren't planning to backport anything to 4.7, but I did add unit testing to verify we don't regress and this happens again. That can be found here: https://github.com/openshift/cluster-ingress-operator/pull/715. Correction: We aren't planning to backport to 4.6 or older. It already is fixed in 4.7; therefore 4.7 doesn't need a backport of anything. Closing this bug as there it is confirmed fixed. The linked PR is for internal unit testing. Opening back up as I just learned we need a BZ open to merge our unit tests https://github.com/openshift/cluster-ingress-operator/pull/715. As of writing, in the latest 4.7 version the env variable sorting issue is no more noted (tested with 4.7.44). The current PR merge does not require further testing from QE as it basically adds a check to verify the sorting for the test cases. Hence marking this as "verified". 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 (Important: OpenShift Container Platform 4.11.0 bug fix and security 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/RHSA-2022:5069 |