This task is for we address the suggestion made in operator-framework/operator-sdk#3499 (comment) for all tests/e2e-tests.

Edit: suggestion is to prefer Expect(len(podNames)).To(Equal(1)) to len(podNames) == 1

Feature Request

Is your feature request related to a problem? Please describe:

We can add a new flag such as host-network. If we run install.sh --host-network, install.sh will open hostNetwork for chaos-daemon.

hostNetwork is also defined in [value.yaml](https://githu

The UI uses the python boto3 library, which doesn't respect the AWS_ENDPOINT environment variable ( boto/boto3#1375 ).

Two updates need to be made, at:
https://github.com/zalando/postgres-operator/blob/master/ui/operator_ui/spiloutils.py#L269
https://github.com/zalando/postgres-operator/blob/master/ui/operator_ui/spiloutils.py#L290

client('s3', endpoint_url=e

Is your feature request related to a problem? Please describe.
Investigate github.com/google/go-containerregistry to replace the heroku/docker-registry-client which seems to be unmaintained.

Describe the solution you'd like
Check the go-containerregistry library and see if it fits our needs.

Describe alternatives you've considered
We already use a forked version of the docker-

Background

As discussed in #223 comments (zalando-incubator/kopf#223 (comment) and below), there is one type of handlers is missing: when an operator restarts, it should process all the objects that existed before, but were neither updated, nor deleted.

Originally, this was a technical solution for @kopf.on.resume() handlers in #96. But then it was

What would you like to be added:
Currently, kubectl kudo version only prints the current version of the KUDO CLI. It should also print the version of the KUDO controller if the con

Feature Request

Is your feature request related to a problem? Please describe:

The shared informer factory has called the WaitForCacheSync method, and it is no longer need to call it separately for each controller. We can remove Informer ().HasSynced because it will not be used.

Given the following CR:

apiVersion: jaegertracing.io/v1
kind: Jaeger
metadata:
  name: simplest
spec:
  allInOne:
    options:
      reporter.grpc.host-port: "opentelemetrycollector-sample-collector-headless.default.svc:14250"

The Jaeger Operator should not be adding the other inferred gRPC options. Using the CR above, the following is the set of options passed to the

We use the adoptopenjdk/openjdk11:slim image, so we should mention this in the OLM CSV.

There is no reason stated in the kubernetes event here:

2s          Warning   CRValidationFailure   Jenkins   Base CR validation failed

The logs:

2019-10-22T09:04:47.894Z        WARN    controller-jenkins      base/validate.go:244    Not found volume for 'backup-script' volume mount in container 'backup' {"cr": "infrastructure", "container": "backup"}
2019-10-22T09:04:47.894Z 

Add example manifests for different persistence options (EBS volume / SSD mount / RAM disk)