Error ""Cannot Terminate — NOT_INSTANTIATED"" observed when network function enters not instantiated state after reset during upgrade failure

Products

• VMware Telco Cloud Automation

Issue/Introduction

• During a Cloud Native Function (CNF) lifecycle management operation, a CNF upgrade fails with the exact error: CNF135027: pre-upgrade hooks failed: timed out waiting for the condition.
• Following this timeout, the Network Function (NF) enters a FAILED_TEMP state.
• After a manual Reset operation is executed by the user, the NF transitions to an inconsistent "Not Instantiated" state, returning a "Cannot Terminate — NOT_INSTANTIATED" error on subsequent lifecycle management (LCM) attempts.

Environment

TCA 2.x

Cause

• The NF enters the unrecoverable "Not Instantiated" state because a manual reset operation was executed while the NF was in the FAILED_TEMP state, which bypassed the required rollback procedures.
• This condition has forced the TCA-M state machine into an inconsistent terminal state.
• Reset was done instead of rollback when the network function is in failed state. 

Resolution

1. Delete the associated namespace from the target workload cluster using the below command to recover the currently affected network function.
   
   kubectl delete ns <namespace-name>
   
   
2. Re-instantiate the Network Function via the TCA-M UI or API using the original CSAR package.
   
   When the network function enter the  FAILED_TEMP state during life cycle management operations, do not execute reset, instead use rollback option. This will safely restore the CNF to its previous working state, which will gracefully exit the FAILED_TEMP lock and unblock subsequent LCM operations.