.. _resilience_upgrade:

******************
Resilience Upgrade
******************

.. important:: When you upgrade a node of your cluster, you must also upgrade the other so that
   they both are running the same version of XiVO. Otherwise, the replication might not work
   properly.

Here are the steps for upgrading a cluster, i.e. two XiVO with :ref:`resilience_architecture`:

#. On the master : deactivate the database replication by commenting the cron in
   :file:`/etc/cron.d/xivo-ha-master`
#. On the slave, deactivate the xivo-check-master-status script cronjob by commenting the line in
   :file:`/etc/cron.d/xivo-ha-slave`
#. On the slave, start the upgrade::

    xivo-slave:~$ xivo-upgrade

#. When the slave has finished, start the upgrade on the master::

    xivo-master:~$ xivo-upgrade

#. When done, launch the database replication manually::

    xivo-master:~$ xivo-master-slave-db-replication <slave ip>

#. Reactivate the cronjobs (see steps 1 and 2)