Appendix: Switching Between DatabasesΒΆ

The /opt/scyld/clusterware/bin/headctl tool performs the database conversion. We suggest taking a snapshot of the virtual machine prior to performing these operations, and preferably making a database backup with managedb as well.

Convert a non-etcd head node to use etcd by executing:

sudo /opt/scyld/clusterware/bin/headctl --use-etcd

which performs a series of steps:

  1. Install the clusterware-etcd package, if not already installed.

  2. Stop the clusterware service.

  3. Export the database to a temporary file.

  4. Toggle the base.ini plugins.database option.

  5. Purge any existing database.

  6. Load the exported database from the temporary file.

  7. Update the firewall for etcd.

  8. Restart the clusterware service.

Once these steps complete, the head node will resume normal operations.

For a multihead cluster this same command should be performed on each head node in turn. Note that the database conversion action detaches that head node from any other cooperating head nodes. When all the head nodes have been converted, you can pick one and join the others to it.

Confirm the head nodes are again working together by executing the following on each head node:

scyld-nodectl status
sudo /opt/scyld/clusterware/bin/managedb --heads

and verify that scyld-nodectl and managedb agree.

When you sure that everything is working as expected, then on each head node remove the clusterware-couchbase package and delete /opt/couchbase.