Description of problem:
On an OCP upgrade from 4.7 -> 4.8, noticed that one of the worker nodes was stuck at "Ready, SchedulingDiabled". On looking at the machine config daemon logs, saw a lot of:
E0330 12:39:45.132218 167828 daemon.go:329] error when evicting pods/"image-registry-cfb9c9c7c-2sq54" -n "openshift-image-registry" (will retry after 5s): Cannot evict pod as it would violate the pod's disruption budget.
Version-Release number of selected component (if applicable):
Reproducible all the time on a cluster with a single image registry replica
On talking to the image-registry folks, it looks like the recent change to add PDB to the image registry deployment caused this issue and this case needs to be handled for non HA image registry.
On the libvirt deploys, there is no storage and we use emptyDir. so this causes the issue.
We're asking the following questions to evaluate whether or not this bug warrants blocking an upgrade edge from either the previous X.Y or X.Y.Z. The ultimate goal is to avoid delivering an update which introduces new risk or reduces cluster functionality in any way. Sample answers are provided to give more context and the UpgradeBlocker flag has been added to this bug. It will be removed if the assessment indicates that this should not block upgrade edges. The expectation is that the assignee answers these questions.
Who is impacted? If we have to block upgrade edges based on this issue, which edges would need blocking?
example: Customers upgrading from 4.y.Z to 4.y+1.z running on GCP with thousands of namespaces, approximately 5% of the subscribed fleet
example: All customers upgrading from 4.y.z to 4.y+1.z fail approximately 10% of the time
What is the impact? Is it serious enough to warrant blocking edges?
example: Up to 2 minute disruption in edge routing
example: Up to 90seconds of API downtime
example: etcd loses quorum and you have to restore from backup
How involved is remediation (even moderately serious impacts might be acceptable if they are easy to mitigate)?
example: Issue resolves itself after five minutes
example: Admin uses oc to fix things
example: Admin must SSH to hosts, restore from backups, or other non standard admin activities
Is this a regression (if all previous versions were also vulnerable, updating to the new, vulnerable version does not increase exposure)?
example: No, it’s always been like this we just never noticed
example: Yes, from 4.y.z to 4.y+1.z Or 4.y.z to 4.y.z+1
Increasing the bugs severity to urgent as we need to know if can impact 4.6.z update edges to 4.7.z.
My understanding is this :
This bug only affects upgrades to 4.8 because of a recent change: https://github.com/openshift/cluster-image-registry-operator/pull/671
From this: https://github.com/openshift/cluster-image-registry-operator/blob/master/pkg/storage/storage.go#L140 the affected platforms will be anything other than AWS,GCP and Azure.
The workaround would be to edit the configs.imageregistry and increase replicas to 2 before starting the upgrade.
Based on comment 3, I'm going to drop UpgradeBlocker and add blocker+.
If this doesn't seem like it will get fixed before 4.8 GAs and folks are ok with that, move to blocker+, restore UpgradeBlocker, and go into more detail about the the "What is the impact?" answer so we can judge blocker-ness.
Adding the relevant test-case string so Sippy associates this bug with the failure, although the test-case is very broad and could fail for many orthogonal reasons.
*** Bug 1944768 has been marked as a duplicate of this bug. ***
PDBs for the registry are very new (bug 1939731), so no impact on 4.7 or earlier.
*** Bug 1945458 has been marked as a duplicate of this bug. ***
tested upgrades with the latest nightlies and works as expected.
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 (Moderate: OpenShift Container Platform 4.8.2 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.