-
Notifications
You must be signed in to change notification settings - Fork 551
fix 1677351. Restrict CLO to single namespace and allow Jaeger to reuse EO without installing CL #85
fix 1677351. Restrict CLO to single namespace and allow Jaeger to reuse EO without installing CL #85
Conversation
|
Hi @jcantrill. Thanks for your PR. I'm waiting for a operator-framework member to verify that this patch is reasonable to test. If it is, they should reply with Once the patch is verified, the new status will be reflected by the I understand the commands that are listed here. DetailsInstructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository. |
|
Any reason we need to wait until post Feb 22 to merge this? Functionally it should be the same aside from requiring users to explicitly install EO and CLO cc @anpingli |
|
looks correct to me based on our earlier conversation |
|
Does CLO require EO to work? Is there a requirement that EO be installed before CLO now that they are separated? |
|
@SamiSousa EO requires either CLO or the Jaeger operator so that required secrets for the resulting deployments created by EO are available, as well as being responsible for creating the CR that the EO watches. We want EO installed before so that we do not run into the dependency resolving issue where CLO is restricted to a single namespace and then EO is installed in the same way (into a single namespace). |
a1c5a3c to
42eda35
Compare
|
|
||
| Once installed, the operator provides the following features: | ||
| * **Create/Destroy**: Deploy an Elasticsearch cluster to the same namespace in which the Elasticsearch custom resource is created. | ||
|
|
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Are there any docs for using this operator with CLO or jaeger? Could these be included either in this description or the description of CLO/jaeger?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I'm not sure what additional documentation applies other then:
This operator only supports OKD Cluster Logging and Jaeger. It is tightly coupled to each and is not currently capable of
being used as a general purpose manager of Elasticsearch clusters running on OKD.
What more is there to say? The operator will act upon any elasticsearch that is created. It is immaterially that CLO or Jaeger created the resource.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
That verbiage in the description is sufficient. I thought there might have been a more complicated setup, but as long as it's clear in the EO description that it is meant for CLO and jaeger, and in the CLO description that EO is required for some functionality, then it's fine.
...ty-operators/elasticsearch-operator/elasticsearch-operator.v0.0.1.clusterserviceversion.yaml
Outdated
Show resolved
Hide resolved
community-operators/cluster-logging/clusterlogging.v0.0.1.clusterserviceversion.yaml
Show resolved
Hide resolved
42eda35 to
d2307e0
Compare
...ty-operators/elasticsearch-operator/elasticsearch-operator.v0.0.1.clusterserviceversion.yaml
Outdated
Show resolved
Hide resolved
e412f19 to
6ec8184
Compare
d9934d4 to
49b8812
Compare
83c427e to
d936566
Compare
|
|
||
| Once installed, the Cluster Logging Operator provides the following features: | ||
| * **Create/Destroy**: Launch and create an aggregated logging stack in the `openshift-logging` namespace. | ||
| * **Create/Destroy**: Launch and create an aggregated logging stack to support the entire OKD cluster. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
If the operator only supports SingleNamespace, how is it supposed to support "the entire OKD cluster"? Unless the application pod is responsible for that.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
The logging stack supports the entire OKD cluster. The cluster-logging-operator only runs in a single namespace.
d936566 to
e7edf69
Compare
|
What's the holdup here? |
e7edf69 to
292fa9a
Compare
…se EO without installing CL bug 1684329. Include openshift-logging in CSV example bug 1683701. CLO ability to list ClusterLogging
292fa9a to
38f204a
Compare
SamiSousa
left a comment
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
/lgtm
|
Operator bundles for |
|
What is the process for getting updates to community operators into a 4.0 cluster? How long does it take from when a PR merges to when I can install a 4.0 cluster with the fixes in it? |
|
We sync with Quay every hour, so you should see updates in that time frame. Please ensure your cluster is running with operator-framework/operator-marketplace#125 which fixes a bug with updates. |
How do I ensure that? If I create a cluster now, will it have that fix? |
|
If you use the latest installer built from master, it should have that fix. You can also confirm by executing |
This PR fixes https://bugzilla.redhat.com/show_bug.cgi?id=1677351 by:
cc @ewolinetz @ecordell @pavolloffay