Switching Stronghold from EE to CSE

Stronghold EE can be upgraded to Stronghold CSE in one of the following ways:

• in a standalone installation;
• in a Deckhouse Platform installation.

Upgrading a standalone installation

Before starting the upgrade, do the following:

1. Check the current Stronghold version in the Version field:

   d8 stronghold status

2. Save the unseal keys and the root token to a secure storage.

3. Create a snapshot of the Stronghold cluster. Example:

   d8 stronghold operator raft snapshot save stronghold-$(date +%F_%H-%M).snap

   Check that the snapshot has been created:

   ls -lh ./stronghold-*.snap

   Store the resulting files outside the Stronghold cluster and outside the nodes it runs on.

4. Prepare the Stronghold CSE 1.16.0 package or binary on each node.

Upgrading a single-node cluster

To upgrade a single-node cluster, do the following:

1. Check the cluster state before the upgrade:

   d8 stronghold status

   The following values must be present:

   Initialized             true
   Sealed                  false
   Version                 1.15.0+ee
   HA Mode                 active
   

2. Stop the stronghold service:

   sudo systemctl stop stronghold

3. Replace the Stronghold EE binary with the Stronghold CSE 1.16.0 binary.

4. Make sure the binary permissions have not changed, or run:

   sudo chmod 0755 /opt/stronghold/stronghold
   sudo chown stronghold:stronghold /opt/stronghold/stronghold

5. Start the stronghold service:

   sudo systemctl start stronghold
   sudo systemctl status stronghold --no-pager

6. Check that the node has started correctly:

   d8 stronghold status

   Check that:

   • the stronghold service is in the Running state;
   • there are no startup errors;
   • the version is 1.16.0 (CSE).

7. If necessary, unseal Stronghold.

Upgrading a multi-node cluster

A multi-node cluster is upgraded node by node so that the cluster remains available.

1. Determine the current leader node:

   d8 stronghold status

   The output contains the address of the current leader node in the HA Cluster field:

   ...
   HA Cluster              https://10.241.32.36:8201
   ...
   

   Note the leader node.

2. Upgrade all follower nodes one by one. On each follower node:

   • Stop the service:

     sudo systemctl stop stronghold

   • Replace the Stronghold EE binary with Stronghold CSE 1.16.0. If necessary, restore the permissions:

     sudo chmod 0755 /opt/stronghold/stronghold
     sudo chown stronghold:stronghold /opt/stronghold/stronghold

   • Start the service:

     sudo systemctl start stronghold
     sudo systemctl status stronghold --no-pager

   • If necessary, unseal the Stronghold node.

   • Check the node state:

     d8 stronghold status

   • Check that:

     • the stronghold service is in the Running state;
     • there are no startup errors;
     • the version is 1.16.0 (CSE).

   • Move on to the next follower node only after the current one has been checked successfully.

3. After upgrading all follower nodes, change the leader node:

   d8 stronghold operator step-down
   # Wait for a while.
   d8 stronghold status

   Expected result: the leader node has changed, and the HA Cluster field shows a different value.

4. Upgrade the former leader node:

   sudo systemctl stop stronghold
   # Replace the Stronghold EE binary with Stronghold CSE 1.16.0.
   # If necessary, restore the permissions.
   # sudo chmod 0755 /opt/stronghold/stronghold
   # sudo chown stronghold:stronghold /opt/stronghold/stronghold
   sudo systemctl start stronghold
   sudo systemctl status stronghold --no-pager
   d8 stronghold status

   Check that:

   • the stronghold service is in the Running state;
   • there are no startup errors;
   • the version is 1.16.0 (CSE).

5. If necessary, unseal the last Stronghold node.

6. Perform the final cluster check:

   d8 stronghold status

   Check that:

   • all cluster nodes are available;
   • there are no replication errors or quorum problems;
   • version 1.16.0 (CSE) is installed on the nodes.