Rolling Software Upgrade Using the SMI Cluster Manager
The AMF software upgrade or in-service upgrade procedure uses the K8s rolling strategy to upgrade the pod images. The pods of a StatefulSet are upgraded sequentially to ensure that the ongoing process remains unaffected.
Initially, a rolling upgrade on a StatefulSet causes a single pod instance to terminate. A pod with an upgraded image replaces the terminated pod. This process continues until all the replicas of the StatefulSet are upgraded.
The terminating pods exit gracefully after completing all the ongoing processes. Other in-service pods continue to receive and process the traffic to provide a seamless software upgrade.
You can control the software upgrade process through the Ops Center CLI.
Note | Each pod needs a minimum of two replicas for high availability. In a worst-case scenario, the processing capacity of the pod may briefly reduce to 50% while the software update is in progress. |
The following figure illustrates the following:
-
The AMF rolling upgrade for AMF REST endpoint pods (two replicas) on protocol worker nodes
-
Along with AMF service pods (three replicas) on service worker nodes

Important | ETCD v3.5.x does not support in-service downgrade to 3.4.x. If you're downgrading from 2023.04.0 builds to previous releases. perform system mode shutdown before downgrade. |