Add support for cluster-scoped CRDs for kube-state-metrics

[raffaelespazzoli]

Signed-off-by: Raffaele Spazzoli [email protected]

this should fix #2707
it introspects the passed gvks to see if they are namespaced. Logic remains the same for namespaced ones, for cluster scoped CRDs, a different watch is built to list/watch at the cluster level.

@camilamacedo86

Description of the change:

it introspects the passed gvks to see if they are namespaced. Logic remains the same for namespaced ones, for cluster scoped CRDs, a different watch is built to list/watch at the cluster level.

Motivation for the change:

enable to produce metrics for cluster scoped CRDs.

[raffaelespazzoli]

@joelanford I don't see the checks making progress, could it be that your renaming the issue has broken the integration with travis-ci?

[raffaelespazzoli]

@joelanford it has happened again: the task: continuous-integration/travis-ci seems to be hung and I can't follow the link to it to check what is going on.
is there a way to relaunch all the checks?

[camilamacedo86]

try to close and re-open the pr @raffaelespazzoli
usually it works.

[joelanford]

Just a few more suggestions.

Can you also add a few lines to the CHANGELOG:

• Under "Deprecated":

  Deprecated `pkg/kubemetrics.NewMetricsStores`. Use `pkg/kubemetrics.NewNamespacedMetricsStores` instead. ([#2809](https://github.com/operator-framework/operator-sdk/pull/2809))

• Under "Added":

  Added support for generating kube-state-metrics metrics for cluster-scoped resources. Also added `pkg/kubemetrics.NewNamespacedMetricsStores` and `pkg/kubemetrics.NewClusterScopedMetricsStores` to support this new feature. ([#2809](https://github.com/operator-framework/operator-sdk/pull/2809))

Looks good to me after addressing these suggestions. Thanks so much for contributing this fix!

[raffaelespazzoli]

@joelanford
what's the policy on error and log messages. It keep failing on those things. and I can't find any mention on this in the style guide.

Checking format of error and log messages...
Error messages do not begin with lower case:
/home/travis/gopath/src/github.com/operator-framework/operator-sdk/pkg/kube-metrics/metrics.go:142:	return false, errors.New("Unable to find type: " + gvk.String() + " in server")
Makefile:209: recipe for target 'test-sanity' failed
make: *** [test-sanity] Error 255

[camilamacedo86]

We just need to fix the changelog.
The links changed and then it will not pass in the CI because they are broken.

[camilamacedo86]

Suggested change

- Added the [`generate csv --deploy-dir --apis-dir --crd-dir`](doc/cli/operator-sdk_generate_csv.md#options) flags to allow configuring input locations for operator manifests and API types directories to the CSV generator in lieu of a config. See the CLI reference doc or `generate csv -h` help text for more details. ([#2511](https://github.com/operator-framework/operator-sdk/pull/2511))

- Added the [`generate csv --deploy-dir --apis-dir --crd-dir`](website/content/en/docs/cli/operator-sdk_generate_csv.md#options) flags to allow configuring input locations for operator manifests and API types directories to the CSV generator in lieu of a config. See the CLI reference doc or `generate csv -h` help text for more details. ([#2511](https://github.com/operator-framework/operator-sdk/pull/2511))

[raffaelespazzoli]

I give up, now it's asking to have error message with lower case:

Checking format of error and log messages...
Error messages do not begin with lower case:
/home/travis/gopath/src/github.com/operator-framework/operator-sdk/pkg/kube-metrics/metrics.go:142:	return false, errors.New("Unable to find type: " + gvk.String() + " in server")
Makefile:209: recipe for target 'test-sanity' failed
make: *** [test-sanity] Error 255
The command "make test-sanity" exited with 2.

you have to fix your sanity check because it's driving me insane :)

[camilamacedo86]

Tested Locally with CRD and Cluster-Scope:

$ kubectl logs deployment.apps/memcached-operator -n memcached
{"level":"info","ts":1586476933.6907554,"logger":"cmd","msg":"Operator Version: 0.0.1"}
{"level":"info","ts":1586476933.6909409,"logger":"cmd","msg":"Go Version: go1.13.3"}
{"level":"info","ts":1586476933.6910954,"logger":"cmd","msg":"Go OS/Arch: linux/amd64"}
{"level":"info","ts":1586476933.691209,"logger":"cmd","msg":"Version of operator-sdk: v0.16.0+git"}
{"level":"info","ts":1586476933.691481,"logger":"leader","msg":"Trying to become the leader."}
{"level":"info","ts":1586476934.0657084,"logger":"leader","msg":"No pre-existing lock was found."}
{"level":"info","ts":1586476934.0715778,"logger":"leader","msg":"Became the leader."}
{"level":"info","ts":1586476934.430387,"logger":"controller-runtime.metrics","msg":"metrics server is starting to listen","addr":"0.0.0.0:8383"}
{"level":"info","ts":1586476934.4311326,"logger":"cmd","msg":"Registering Components."}
{"level":"info","ts":1586476935.1817207,"logger":"metrics","msg":"Metrics Service object created","Service.Name":"memcached-operator-metrics","Service.Namespace":"memcached"}
{"level":"info","ts":1586476935.5474622,"logger":"cmd","msg":"Starting the Cmd."}
{"level":"info","ts":1586476935.548912,"logger":"controller-runtime.manager","msg":"starting metrics server","path":"/metrics"}
{"level":"info","ts":1586476935.548941,"logger":"controller-runtime.controller","msg":"Starting EventSource","controller":"memcached-controller","source":"kind source: /, Kind="}
{"level":"info","ts":1586476935.6514418,"logger":"controller-runtime.controller","msg":"Starting EventSource","controller":"memcached-controller","source":"kind source: /, Kind="}
{"level":"info","ts":1586476935.7533202,"logger":"controller-runtime.controller","msg":"Starting EventSource","controller":"memcached-controller","source":"kind source: /, Kind="}
{"level":"info","ts":1586476935.8547926,"logger":"controller-runtime.controller","msg":"Starting Controller","controller":"memcached-controller"}
{"level":"info","ts":1586476935.8550828,"logger":"controller-runtime.controller","msg":"Starting workers","controller":"memcached-controller","worker count":1}
{"level":"info","ts":1586476935.8556492,"logger":"controller_memcached","msg":"Reconciling Memcached.","Request.Namespace":"","Request.Name":"example-memcached"}

We just need to fix the rebase/conflict in the changelog to pass in the CI.
Great work 👍

/lgtm
/approve

[openshift-ci-robot]

New changes are detected. LGTM label has been removed.