Skip to content

ClusterRoleBinding applied asynchronously leading to topology provider failures #773

@NickLarsenNZ

Description

@NickLarsenNZ

Affected Stackable version

All up until and including SDP 26.3.0

Affected Apache HDFS version

N/A

Current and expected behavior

When the Topology Provider runs, it is assumed that the necessary permissions are there for the Kubernetes API.

There is a small window for a new HdfsCluster's where the ClusterRoleBinding hasn't been applied by the time the Topology Provider starts.

Possible solution

Instead of concurrently applying ClusterRoleBindings and StatefulSets, do them in sequence.

The overhead should be negligible.

Additional context

No response

Environment

No response

Would you like to work on fixing this bug?

yes

Metadata

Metadata

Assignees

No one assigned

    Labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions