NiFi instance did not start

[deebify]

I tried to install demo from stackablectl demo using
stackablectl demo install nifi-kafka-druid-water-level-data
NiFi is not deployed.

[INFO ] Installing demo nifi-kafka-druid-water-level-data
[INFO ] Installing stack nifi-kafka-druid-superset-s3
[INFO ] Installing release 23.1
[INFO ] Installing commons operator in version 23.1.0
[INFO ] The release commons-operator in version 23.1.0 is already installed, not installing it
[INFO ] Installing druid operator in version 23.1.0
[INFO ] The release druid-operator in version 23.1.0 is already installed, not installing it
[INFO ] Installing kafka operator in version 23.1.0
[INFO ] The release kafka-operator in version 23.1.0 is already installed, not installing it
[INFO ] Installing nifi operator in version 23.1.0
[INFO ] The release nifi-operator in version 23.1.0 is already installed, not installing it
[INFO ] Installing secret operator in version 23.1.0
[INFO ] The release secret-operator in version 23.1.0 is already installed, not installing it
[INFO ] Installing superset operator in version 23.1.0
[INFO ] The release superset-operator in version 23.1.0 is already installed, not installing it
[INFO ] Installing zookeeper operator in version 23.1.0
[INFO ] The release zookeeper-operator in version 23.1.0 is already installed, not installing it
[INFO ] Installing components of stack nifi-kafka-druid-superset-s3
[INFO ] The release minio-druid in version 4.0.15 is already installed, not installing it
[INFO ] The release postgresql-druid in version 12.1.5 is already installed, not installing it
[INFO ] The release postgresql-superset in version 12.1.5 is already installed, not installing it
[ERROR] failed with status 500 Internal Server Error
[ERROR] ApiError: failed to create typed patch object (/nifi; nifi.stackable.tech/v1alpha1, Kind=NifiCluster): .spec.zookeeperConfigMapName: field not declared in schema: (ErrorResponse { status: "Failure", message: "failed to create typed patch object (/nifi; nifi.stackable.tech/v1alpha1, Kind=NifiCluster): .spec.zookeeperConfigMapName: field not declared in schema", reason: "", code: 500 })

I tried to get logs from stackable NiFi operator, what I see
Error("missing field zookeeperConfigMapName", line: 1, column: 4827)

Stackable CLI version
╰─$ stackablectl -V 255 ↵ stackablectl 0.8.0

Stackable operators installed
╰─$ stackablectl operator installed 2 ↵ OPERATOR VERSION NAMESPACE STATUS LAST UPDATED commons 23.1.0 default deployed 2023-03-03 01:07:41.297103143 +0200 EET druid 23.1.0 default deployed 2023-03-03 01:39:09.322565151 +0200 EET kafka 23.1.0 default deployed 2023-03-03 01:39:12.304941504 +0200 EET nifi 23.1.0 default deployed 2023-03-03 01:42:31.01013892 +0200 EET secret 23.1.0 default deployed 2023-03-03 01:07:44.256278371 +0200 EET superset 23.1.0 default deployed 2023-03-03 01:39:15.352646993 +0200 EET zookeeper 23.1.0 default deployed 2023-03-03 01:07:45.235510534 +0200 EET

Here are logs from the stackable NiFi operator pod:
2023-03-02T23:42:31.938062Z INFO stackable_operator::utils: Starting Stackable Operator for Apache NiFi 2023-03-02T23:42:31.938080Z INFO stackable_operator::utils: This is version 23.1.0 (Git information: 23.1.0), built for x86_64-unknown-linux-gnu by rustc 1.65.0 (897e37553 2022-11-02) at Mon, 23 Jan 2023 11:14:51 +0000 2023-03-02T23:42:31.960351Z INFO kube_runtime::controller: press ctrl+c to shut down gracefully Error("missing field zookeeperConfigMapName", line: 1, column: 4827) 2023-03-02T23:42:44.830591Z ERROR stackable_operator::logging::controller: Failed to reconcile object controller.name="nificluster.nifi.stackable.tech" error=event queue error error.sources=[failed to perform initial object list: Error deserializing response, Error deserializing response, missing field zookeeperConfigMapName at line 1 column 4827]

[sbernauer]

Hi @deebify, thanks for reporting this!

I did try to reproduce but was not able to do so, the demo seems to work fine.
In our currently nightly version we have moved some fields (e.g. zookeeperConfigMapName) from spec.zookeeperConfigMapName to spec.clusterConfig.zookeeperConfigMapName. So the CustomResourceDefinition (CRD) of the nightly version does not have the field spec.zookeeperConfigMapName any more.
The demos however always run on a stable version, 23.1 in this case. The demo does use the correct field for the 23.1 release here.
The only remaining option is that you have a different version of the CRD installed. Could you please delete the CRD with kubectl delete crd nificlusters.nifi.stackable.tech and re-create it in the correct version with kubectl apply -f https://raw.githubusercontent.com/stackabletech/nifi-operator/23.1.0/deploy/helm/nifi-operator/crds/crds.yaml. Afterwards you should be able to delete the Nifi cluster (kubectl delete nifi nifi) and re-run the demo install statement.

A solution that should work in any case is to delete your k8s and start with a fresh k8s.

Hope that helps,
Sebastian