Ondat Upgrade

Overview

This guide provides instructions on how to upgrade Ondat.

Upgrading An Ondat `v2` Cluster

Prerequisites

⚠️ Ensure that you have read the PIDs prerequisite introduced in Ondat v2.3 and that you have checked the init container logs to ensure your environments PID limits are set correctly.

💡 Pull down the new Ondat container image storageos/node:v2.7.0 onto the nodes beforehand so that the cluster spins up faster.

💡 Speak with our support team here so we can assist you with your upgrade.

⚠️ If you are upgrading to v2.7.0, you will only be able to downgrade to v2.6.0 due to the mapping changes made in the Data Plane. For more information, review the release notes.

Procedure

Step 1 - Backup Ondat Deployment Manifests

Make sure that you keep a backup of all the Ondat YAML files. You can also backup the StatefulSet yaml files to keep track of the replicas.

kubectl get pod -n storageos-operator -o yaml > storageos_operator.yaml
kubectl get storageoscluster -n storageos-operator -o yaml > storageos_cr.yaml
kubectl get statefulset --all-namespaces > statefulset-sizes.yaml

Step 2 - Generate The `storageos-config.yaml` Manifest

Run the following commands to generate the storageos-config.yaml file that is used to upgrade your cluster based on your current configuration.

SECRET_NAME=$(kubectl get storageoscluster -A -o=jsonpath='{.items[0].spec.secretRefName}')
SECRET_NAMESPACE=$(kubectl get storageoscluster -A -o=jsonpath='{.items[0].spec.secretRefNamespace}')
kubectl get secret -n $SECRET_NAMESPACE $SECRET_NAME -oyaml > /tmp/storageos-config.yaml
echo "---" >> /tmp/storageos-config.yaml
STOS_NAME=$(kubectl get storageoscluster -A -o=jsonpath='{.items[0].metadata.name}')
STOS_NAMESPACE=$(kubectl get storageoscluster -A -o=jsonpath='{.items[0].metadata.namespace}')
kubectl get storageoscluster -n $STOS_NAMESPACE $STOS_NAME -oyaml >> /tmp/storageos-config.yaml

This file will include the manifest of 2 Kubernetes objects, the authentication Secret object and the StorageOS CR object.
Edit StorageOSCluster object in the /tmp/storageos-config.yaml file by removing all images in the image section and modify the storageos/node image to:
```
images:
    nodeContainer: "storageos/node:v2.7.0"
```

Step 3 - Scale Down Stateful Applications To Zero

Scale all of the stateful applications that use Ondat volumes to 0.

Step 4 - Upgrade Ondat

Run the following command to conduct the upgrade:

kubectl storageos upgrade --uninstall-stos-operator-namespace storageos-operator --stos-cluster-yaml /tmp/storageos-config.yaml --etcd-endpoints "<ETCD-IP1>:2379,<ETCD-IP2>:2379,<ETCD-IP3>:2379"

💡 The plugin uses the --uninstall-stos-operator-namespace argument because it uninstalls the cluster first and then reinstalls it with the new version.

The etcd endpoints should correspond to the endpoints you have in the .spec.kvBackend section of StorageOS object in the /tmp/storageos-config.yaml file.

💡 If at any point something goes wrong with the upgrade process, backups of all the relevant Kubernetes manifests can be found in ~/.kube/storageos/.

Step 5 - Verifying The Ondat Upgrade

Run the following commands to inspect Ondat’s resources are back online (the core components should all be in a RUNNING status)
```
kubectl get all --namespace=storageos
kubectl get all --namespace=storageos-etcd
```

Step 6 - Scale Up Stateful Applications

Once the Ondat upgrade is complete and the core components are back online, scale up the stateful applications that use Ondat volumes back up to their respective replica count.