6 Rolling Back OSO

This chapter provides information about rolling back Oracle Communications Cloud Native Core, Operations Services Overlay (OSO) deployment to previous releases.

6.1 Supported Rollback Paths

The following table lists the supported rollback paths for OSO:

Table 6-1 Supported Rollback Paths

Source Release	Target Release
25.2.200	25.2.1xx
25.2.200	25.1.2xx

6.2 Prerequisites

Following are the prerequisites to rollback OSO:

• Ensure that the system has OSO 25.2.200 installed and the OSO is working properly.
• Verify if all the pods and services are up and running.

6.3 Rollback Tasks

To roll back from OSO 25.2.200 to previous releases 25.1.2xx or 25.2.1xx:

1. Run the following command to check the installed Helm charts.

   $ helm -n oso ls

   For example:

   
   NAME    	NAMESPACE	REVISION	UPDATED                                       STATUS  	CHART               	APP VERSION
   oso-a   	oso      	2       	2025-08-04 17:46:51.998024802 +0000 UTC	deployed	alertmanager-1.15.1 	25.2.200-39-gbc755fc
   oso-alr-cfg    oso      	2       	2025-08-04 17:47:18.547249161 +0000 UTC	deployed      oso-alr-config-0.1.0    25.2.200-39-gbc755fc
   oso-apm 	oso      	1       	2025-08-01 22:08:15.58454657  +0000 UTC	deployed      apm-service-25.2.200	25.2.200-alpha.1
   oso-p   	oso      	2       	2025-08-04 17:46:51.276209128 +0000 UTC	deployed      prometheus-27.5.0   	25.2.200-39-gbc755fc

2. Pick a chart to rollback and check the revision you want to roll back your release to. The chart can be either Prometheus or alertmanager.

   $ helm -n <namespace> history <oso-release-name>

   For example:

   $ helm -n oso history oso-p

   Sample output

   
   REVISION        UPDATED                         STATUS          CHART                   APP VERSION             DESCRIPTION
   1               Fri Aug  1 17:46:51 2025        superseded      prometheus-27.5.0       25.2.200   Install complete 
   2               Mon Aug  4 14:46:19 2025        deployed        prometheus-27.5.0       25.2.200   Upgrade complete
   

3. Pick the revision number to which you wish to roll back your release into. In above example, it will be 1. Run the following command to roll back:

   $ helm  -n <namespace> rollback <oso-release-name> <oso revision number>

   For example:

   $ helm -n oso rollback oso 1

   Sample output

   Rollback was a success! Happy Helming!

   For example:

   $ helm -n oso history oso-p 
   REVISION           UPDATED                        STATUS           CHART                 APP VERSION       DESCRIPTION
    1               Fri Aug  1 17:46:51 2025        superseded      prometheus-27.5.0       25.1.200       Install complete
    2               Mon Aug  4 14:46:19 2025        superseded      prometheus-27.5.0       25.2.100       Upgrade complete 
    3               Thu Aug  7 12:37:02 2025        deployed        prometheus-27.5.0       25.1.200       Rollback to 1

4. [

   Note:

   This step is only required for Prometheus chart.
   ] After Helm rollback a new Prometheus pod is created but in pending state. To make it into running state, we have to detach the PVC from old pod and attach to the new pod using below steps:

   Note:

   Find Prometheus deployment name and replace it in the following commands, Both the commands given below should be run at the same time.

   
   $ kubectl get deployment -n <oso-namespace>
   # Find Prometheus deployment name and replace it in below commands, Both the commands below needs to be run at the same time
   $ kubectl -n <oso-namespace> scale deploy <oso-deployment-name> --replicas=0
   $ kubectl -n <oso-namespace> scale deploy <oso-deployment-name> --replicas=1
   

   For example with sample output

   
   $ kubectl -n oso scale deploy oso-p-prom-svr --replicas=0
   deployment.apps/oso-prom-svr scaled 
   $ kubectl -n oso scale deploy oso-p-prom-svr --replicas=1 
   deployment.apps/oso-prom-svr scaled

5. Verify if all the pods are up and running with the latest versions and changes, perform the following command:

   $ kubectl get pods --namespace <namespace>

   For example:

   $ kubectl get pods -n occne-infra

   Sample output

   
   NAME                            READY   STATUS    RESTARTS   AGE
   oso-prom-alm-0                  2/2     Running   0          25h
   oso-prom-alm-1                  2/2     Running   0          25h
   oso-prom-svr-84c4hfd488-qsnvx   2/2     Running   0          14s

6. Verify if the upgraded OSO is working as expected by performing the Verifying Installation procedure.