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
There are many Pravega Spec parameters still remaining which cannot be updated once the pravega cluster has been created. All the spec fields other than the the version, replica count, DNS name, JVM options as well as the pravega options still come under this category. While not allowing updates on some parameters might make sense, it will be necessary to allow updates to a few others including controllerResources, segmentStoreResources, segmentStoreEnvVars, segmentStoreSecret etc.
The operator reconcile logic would need to be modified so that updates to necessary fields are allowed during the lifetime of the cluster.
A maintenance mode can be added to stop operator from watching all resources and allow the user to make any changes to recover the cluster.
The text was updated successfully, but these errors were encountered:
SrishT
changed the title
Many Pravega Spec parameters are still updatable at runtime
Many Pravega Spec parameters are still not updatable at runtime
Jun 10, 2021
Description
There are many Pravega Spec parameters still remaining which cannot be updated once the pravega cluster has been created. All the spec fields other than the the version, replica count, DNS name, JVM options as well as the pravega options still come under this category. While not allowing updates on some parameters might make sense, it will be necessary to allow updates to a few others including controllerResources, segmentStoreResources, segmentStoreEnvVars, segmentStoreSecret etc.
Importance
must have
Location
pkg/controller/pravegacluster/pravegacluster_controller.go
Suggestions for an improvement
The operator reconcile logic would need to be modified so that updates to necessary fields are allowed during the lifetime of the cluster.
A maintenance mode can be added to stop operator from watching all resources and allow the user to make any changes to recover the cluster.
The text was updated successfully, but these errors were encountered: