You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Seems like the issue is with the helm chart as It's not possible to deploy a NifiCluster with only autoscaled node groups. The NifiCluster CRD requires that you specify at least one node in the spec.nodes list.
I disabled the autoscaling in the helm chart and did it manually using separate deployment. It worked flawlessly will orchestrate this using TF
It just occurred to me that i mentioned this as a constraint when i raised the PR to contribute this feature: #89
It's not possible to deploy a NifiCluster with only autoscaled node groups. The NifiCluster CRD requires that you specify at least one node in the spec.nodes list. Do we want to support this? If so, we may need to adjust the cluster initialization logic in the NifiCluster controller.
This could be something we evaluate changing. At the very least this constraint needs to be in the documentation.
What steps will reproduce the bug?
Environment
Google Kubernetes Engine 1.24.7-gke.900
Nifikop 1.0.0
nificluster 1.17.0
Assign node labels
NifiCluster.spec.nodes
What is the expected behavior?
Nifi nodes should autoscale seamlessly.
PS:- I have tried scaling up and down both yields same error.
What do you see instead?
nifikop operator fails with
"msg":"Observed a panic in reconciler: runtime error: invalid memory address or nil pointer dereference"
Possible solution
Seems like the error is stemming from GetCreationTimeOrderedNodes() method.
Error
NiFiKop version
v.1.0.0
Golang version
go.1.19
Kubernetes version
1.24
NiFi version
1.17.0
Additional context
NA
The text was updated successfully, but these errors were encountered: