Replace Failing Front-end

Follow this guide as the oneadmin user. To switch the user run:

sudo -i -u oneadmin

Back up files

This command has to be executed in the failing Host.

Create the backup directory:

BAK_DIR=~/one_backup
mkdir -p $BAK_DIR

Copy configuration directory /etc/one to backup directory:

cp -rp --parents /etc/one $BAK_DIR

Copy remotes directory /var/lib/one/remotes to backup directory:

cp -rp --parents /var/lib/one/remotes $BAK_DIR

Copy keys directory /var/lib/one/.one to backup directory:

cp -rp --parents /var/lib/one/.one $BAK_DIR

Back up the database:

onedb backup -S <database_host> -u <user> -p <password> -d <database_name> -P <port>

Copy the database backup file to the backup directory:

cp -rp <onedb_backup> $BAK_DIR

Remove failing Front-end from HA

Stop all the OpenNebula services systemctl stop opennebula*. This command has to be executed in the failing Host.

systemctl stop opennebula*

Remove it from HA using the CLI command onezone onezone server-del <zone_id> <server_id>. This command has to be executed in the leader.

onezone server-del <zone_id> <server_id>

Add the server back to the HA

To add the a replacement server into the HA cluster, please follow this guide.