Upgrade an Online Docker Cluster, Full Capacity
A Docker cluster with multiple nodes can be upgraded live, while maintaining full capacity.
Understanding the Full Capacity Upgrade
The context and overall requirements for upgrading a live cluster without the introduction of one or more additional nodes are described in Upgrade-Procedure Selection: this explains the node-by-node upgrade of the cluster, using the swap rebalance process to minimize overhead.
It is not possible to upgrade a Docker cluster with a single node. A second node will needed to ensure data is transferred. |
The precise steps for this procedure are detailed on this page, below. A full understanding of the information in Upgrade-Procedure Selection should be acquired, before proceeding. Yous should also be familiar with using Docker commands to create nodes. The procedure assumes that:
-
The cluster to be upgraded must continue to serve data throughout the cluster-upgrade process.
-
A single, additional node can be created to act as a swap node.
-
During the cluster-upgrade process, swap nodes and upgraded nodes will be introduced to the cluster by the node-addition procedure; while failed over spare nodes and nodes to be upgraded will be withdrawn from the cluster by means of the node removal and swap rebalance procedures.
For information on node-addition, see Clusters. For information on node-removal, see Removal. For information on swap rebalance, see Swap Rebalance.
-
Nodes will be upgraded one at a time.
1. Prepare the Cluster
Carry out a full backup of the cluster’s data. This can be done using the REST-API or the Backup Service.
2. Create the Spare Node
Use the standard Docker tools to create your node. (For more information read Install Couchbase Server Using Docker). In this case, you will create a node that can act as cluster manager. For example:
docker run -d --name sparenode -p 28091-28096:8091-8096 -p 11310-11311:11210-11211 couchbase:latest
will create a new node (sparenode
) using the latest Couchbase server image.
Note that the port mappings have been set so as not to clash with the external port settings of the existing node that is managing the cluster.
The node must be created as managing cluster. Later on, the existing managing node will be removed and upgraded, so the new node will be needed to communicate with the cluster. |
3. Add the Spare Node to the Cluster and Rebalance the Cluster
The new node must be added to the cluster, then the cluster must be rebalanced. This can be done through the administration console, or the REST-API. (For information on adding nodes and rebalancing, read Add a Node and Rebalance).
Ensure that the sparenode can be used to access the cluster through the administration console.
It will use the same IP address, but a different port number.
(In the example above, this is defined as 28091 ).
|
4. Remove a Node For Upgrading
Using the sparenode
, remove one of the nodes for upgrading from the cluster and rebalance the cluster.
5. Delete the Node Container
The docker
command is used to delete the node and its associated volume.
For example:
docker rm -f nodename
docker volume prune
6. Upgrade the Node Container
The command used to upgrade the node depends on whether it is a standard node, or the node that used to administer the cluster. When recreating the container for the administration node, the docker
command must set up the ports to match the settings of the node that has been removed.
-
Standard Node
-
Administration Node
The following command will create a new container for the standard node.
docker run -d --name nodename couchbase:latest
The following command will create a new container for the administration node.
docker run -d --name nodename -p 8091-8096:8091-8096 -p 11210-11211:11210-11211 couchbase:latest
Note that the external port addresses match those of the administration node that has been removed.
7. Add the New Node to the Cluster and Rebalance
Use the sparenode
to add the new node to the cluster, then rebalance the cluster.
This can be done through the administration console, or the REST-API.
(For information on adding nodes and rebalancing, read Add a Node and Rebalance).
9. Remove the Spare Node
The upgraded administration node can be used to remove the spare node from the cluster and rebalance the cluster.
Once this has been done, then the docker
command can be used to remove the spare node’s container and volume.
docker rm -f sparenode
docker volume prune
The cluster is now fully upgraded.