Skip to main content
Version: 24.04

Upgrade Centreon HA from Centreon 21.04

This chapter describes how to upgrade your Centreon HA platform from version 21.04 to version 24.04.

You cannot simply upgrade a platform with Centreon HA (or Centreon Failover) from a version earlier than 21.04 to version 24.04, as CentOS 7 is no longer supported. You need to migrate your platform to a supported OS, then install Centreon HA on the new platform. You can also contact Centreon to order a migration service.

Prerequisites​

Suspend cluster resources management​

In order to avoid a failover of the cluster during the update, it is necessary to unmanage all Centreon resources, as well as MariaDB.

pcs property set maintenance-mode=true

Perform a backup​

Be sure that you have fully backed up your environment for the following servers:

  • Central server
  • Database server

Update the RPM signing key​

For security reasons, the keys used to sign Centreon RPMs are rotated regularly. The last change occurred on October 14, 2021. When upgrading from an older version, you need to go through the key rotation procedure, to remove the old key and install the new one.

Upgrade process​

To perform the upgrade, please follow the official documentation only on the active central node and active database node if needed.

Then perform the following commands only on the Central Servers:

mv /etc/centreon-ha/centreon_central_sync.pm.rpmsave /etc/centreon-ha/centreon_central_sync.pm

On the passive central node, move the "install" directory to avoid getting the "upgrade" screen in the WUI in the event of a further exchange of roles.

mv /usr/share/centreon/www/install /var/lib/centreon/installs/install-update-YYYY-MM-DD
sudo -u apache /usr/share/centreon/bin/console cache:clear

Removing cron jobs​

The RPM upgrade puts cron jobs back in place on the central and database servers. Remove them to avoid concurrent executions: 

rm -rf /etc/cron.d/centreon
rm -rf /etc/cron.d/centstorage
rm -f /etc/cron.d/centreon-ha-mysql

As you have deleted the centreon-ha-mysql cron, check that the following line appears in the server section of the /etc/my.cnf.d/server.cnf file:

expire_logs_days=7

If the line is not there, add it, then restart the ms_mysql resource:

pcs resource restart ms_mysql

Reset the permissions for centreon_central_sync resource​

The RPM upgrade puts the permissions back in place on the central servers. Change them using these commands:

chmod 775 /var/log/centreon-engine/
mkdir /var/log/centreon-engine/archives
chown centreon-engine: /var/log/centreon-engine/archives
chmod 775 /var/log/centreon-engine/archives/
find /var/log/centreon-engine/ -type f -exec chmod 664 {} \;
find /usr/share/centreon/www/img/media -type d -exec chmod 775 {} \;
find /usr/share/centreon/www/img/media -type f \( ! -iname ".keep" ! -iname ".htaccess" \) -exec chmod 664 {} \;

Cluster upgrade​

From Centreon 22.04, MariaDB Replication is based on GTID. It is necessary to destroy the cluster completely and then reconfigure it with the latest version of the Centreon and MariaDB Replication mechanisms.

Maintenance mode and backup​

Perform a backup of the cluster using:

pcs config backup centreon_cluster
pcs config export pcs-commands | sed -e :a -e '/\\$/N; s/\\\n//; ta' | sed 's/-f tmp-cib.xml//' | egrep "create|group" | egrep -v "(mysql|php|cbd_rrd)" > centreon_pcs_command.sh

Check that the file centreon_cluster.tar.bz2 exists before continuing this procedure.

ls -l centreon_cluster.tar.bz2

You should have a result like this:

-rw------- 1 root root 2777 May  3 17:49 centreon_cluster.tar.bz2

Then check the file centreon_pcs_command.sh. The export command may display some warning lines, but this is not blocking.

cat centreon_pcs_command.sh

The content should looks like this:

pcs resource create vip ocf:heartbeat:IPaddr2 broadcast=@VIP_BROADCAST_IPADDR@ cidr_netmask=@VIP_CIDR_NETMASK@ flush_routes=true ip=@VIP_IPADDR@ nic=@VIP_IFNAME@ op monitor interval=10s timeout=20s start interval=0s timeout=20s stop interval=0s timeout=20s meta target-role=started
pcs resource create http systemd:httpd24-httpd op monitor interval=5s timeout=20s start interval=0s timeout=40s stop interval=0s timeout=40s meta target-role=started
pcs resource create gorgone systemd:gorgoned op monitor interval=5s timeout=20s start interval=0s timeout=90s stop interval=0s timeout=90s meta target-role=started
pcs resource create centreon_central_sync systemd:centreon-central-sync op monitor interval=5s timeout=20s start interval=0s timeout=90s stop interval=0s timeout=90s meta target-role=started
pcs resource create cbd_central_broker systemd:cbd-sql op monitor interval=5s timeout=30s start interval=0s timeout=90s stop interval=0s timeout=90s meta target-role=started
pcs resource create centengine systemd:centengine op monitor interval=5s timeout=30s start interval=0s timeout=90s stop interval=0s timeout=90s meta multiple-active=stop_start target-role=started
pcs resource create centreontrapd systemd:centreontrapd op monitor interval=5s timeout=20s start interval=0s timeout=30s stop interval=0s timeout=30s meta target-role=started
pcs resource create snmptrapd systemd:snmptrapd op monitor interval=5s timeout=20s start interval=0s timeout=30s stop interval=0s timeout=30s meta target-role=started
pcs resource group add centreon vip http gorgone centreon_central_sync cbd_central_broker centengine centreontrapd snmptrapd

This file will be necessary to recreate all the resources of your cluster.

Delete the resources​

These commands should run only on the active central node:

pcs resource delete ms_mysql --force
pcs resource delete cbd_rrd --force
pcs resource delete php7 --force
pcs resource delete centreon --force

Reconfigure MariaDB​

It is necessary to modify the mysql configuration by editing the file /etc/my.cnf.d/server.cnf:

On the 2 Central servers in HA 2 nodes On the 2 Database servers in HA 4 nodes.

[server]
...
skip-slave-start
log-slave-updates
gtid_strict_mode=ON
expire_logs_days=7
ignore-db-dir=lost+found
...

Launch GTID replication​

Run this command on the secondary database node:

mysqladmin -p shutdown

It is important to make sure that MariaDB is completely shut down. Run this command and check that it returns no output:

ps -ef | grep mariadb[d]

Once the service is stopped on the secondary database node, run the synchronization script from the primary database node:

mysqladmin -p shutdown
systemctl restart mariadb
/usr/share/centreon-ha/bin/mysql-sync-bigdb.sh

This script's output is very verbose and you can't expect to understand everything, so to make sure it went well, focus on the last lines of its output, checking that it looks like this:

Umount and Delete LVM snapshot
  Logical volume "dbbackupdatadir" successfully removed
Start MySQL Slave
Start Replication
Id  User    Host    db  Command Time    State   Info    Progress
[variable number of lines]

The important thing to check is that Start MySQL Slave and Start Replication are present and are not followed by any errors.

In addition, the output of this command must display only OK results:

/usr/share/centreon-ha/bin/mysql-check-status.sh

The expected output is:

Connection Status '@CENTRAL_MASTER_NAME@' [OK]
Connection Status '@CENTRAL_SLAVE_NAME@' [OK]
Slave Thread Status [OK]
Position Status [OK]

Restart Centreon process​

Then to restart all the processes on the active central node:

systemctl restart cbd-sql cbd gorgoned centengine centreontrapd

And on the passive central node:

systemctl restart cbd

Clean broker memory files​

WARNING: perform this command only on the passive central node.

Before resuming cluster resource management, to avoid broker issues, clean up all the .memory., .unprocessed. or .queue. files:

rm -rf /var/lib/centreon-broker/central-broker-master.memory*
rm -rf /var/lib/centreon-broker/central-broker-master.queue*
rm -rf /var/lib/centreon-broker/central-broker-master.unprocessed*

Recreate the cluster resources​

To be run only on one central node:

WARNING: the syntax of the following command depends on the Linux Distribution you are using.

You can find @CENTRAL_MASTER_NAME@ @CENTRAL_SLAVE_NAME@ @MARIADB_REPL_USER@ @MARIADB_REPL_USER@ variable in /etc/centreon-ha/mysql-resources.sh.

pcs resource create "ms_mysql" \
    ocf:heartbeat:mariadb-centreon \
    config="/etc/my.cnf.d/server.cnf" \
    pid="/var/lib/mysql/mysql.pid" \
    datadir="/var/lib/mysql" \
    socket="/var/lib/mysql/mysql.sock" \
    binary="/usr/bin/mysqld_safe" \
    node_list="@CENTRAL_MASTER_NAME@ @CENTRAL_SLAVE_NAME@" \
    replication_user="@MARIADB_REPL_USER@" \
    replication_passwd='@MARIADB_REPL_PASSWD@' \
    test_user="@MARIADB_REPL_USER@" \
    test_passwd="@MARIADB_REPL_PASSWD@" \
    test_table='centreon.host'

WARNING: the syntax of the following command depends on the Linux Distribution you are using.

pcs resource promotable ms_mysql \
    master-node-max="1" \
    clone_max="2" \
    globally-unique="false" \
    clone-node-max="1" \
    notify="true"

PHP resource​

pcs resource create "php" \
    systemd:php-fpm \
    meta target-role="started" \
    op start interval="0s" timeout="30s" \
    stop interval="0s" timeout="30s" \
    monitor interval="5s" timeout="30s" \
    clone

RRD broker resource​

pcs resource create "cbd_rrd" \
    systemd:cbd \
    meta target-role="started" \
    op start interval="0s" timeout="90s" \
    stop interval="0s" timeout="90s" \
    monitor interval="20s" timeout="30s" \
    clone

Recreating the centreon resource group​

bash centreon_pcs_command.sh

Recreating the constraint​

pcs constraint colocation add master "ms_mysql-clone" with "centreon"
pcs constraint order stop centreon then demote ms_mysql-clone

Resuming cluster resource management​

Now that the update is finished, the resources can be managed again:

pcs property set maintenance-mode=false
pcs resource cleanup

Check the health of the cluster​

You can monitor the cluster's resources in real time using the crm_mon -fr command:

INFO: The -fr option allows you to display all resources even if they are disabled.

Stack: corosync
Current DC: @CENTRAL_SLAVE_NAME@ (version 1.1.20-5.el7_7.2-3c4c782f70) - partition with quorum
Last updated: Thu Feb 20 13:14:17 2020
Last change: Thu Feb 20 09:25:54 2020 by root via crm_attribute on @CENTRAL_MASTER_NAME@

2 nodes configured
14 resources configured

Online: [ @CENTRAL_MASTER_NAME@ @CENTRAL_SLAVE_NAME@ ]

Active resources:

 Master/Slave Set: ms_mysql-master [ms_mysql]
     Masters: [ @CENTRAL_MASTER_NAME@ ]
     Slaves: [ @CENTRAL_SLAVE_NAME@ ]
 Clone Set: cbd_rrd-clone [cbd_rrd]
     Started: [ @CENTRAL_MASTER_NAME@ @CENTRAL_SLAVE_NAME@ ]
 Resource Group: centreon
     vip        (ocf::heartbeat:IPaddr2):   Started @CENTRAL_MASTER_NAME@
     http   (systemd:httpd24-httpd):        Started @CENTRAL_MASTER_NAME@
     gorgone    (systemd:gorgoned):     Started @CENTRAL_MASTER_NAME@
     centreon_central_sync  (systemd:centreon-central-sync):        Started @CENTRAL_MASTER_NAME@
     centreontrapd  (systemd:centreontrapd):        Started @CENTRAL_MASTER_NAME@
     snmptrapd  (systemd:snmptrapd):    Started @CENTRAL_MASTER_NAME@
     cbd_central_broker (systemd:cbd-sql):  Started @CENTRAL_MASTER_NAME@
     centengine (systemd:centengine):   Started @CENTRAL_MASTER_NAME@
 Clone Set: php-clone [php]
     Started: [ @CENTRAL_MASTER_NAME@ @CENTRAL_SLAVE_NAME@ ]

Disabled resources​

When you do a crm_mon -fr and you have a resource that is disabled:

...
 Master/Slave Set: ms_mysql-master [ms_mysql]
     Masters: [ @DATABASE_MASTER_NAME@ ]
     Slaves: [ @DATABASE_SLAVE_NAME@ ]
     Stopped: [ @CENTRAL_MASTER_NAME@ @CENTRAL_SLAVE_NAME@ ]
vip_mysql       (ocf::heartbeat:IPaddr2):       Stopped (disabled)
...

You must enable the resource with the following command:

pcs resource enable @RESSOURCE_NAME@

In our case:

pcs resource enable vip_mysql

Verifying platform stability​

You should now check that everything works fine:

  • Access to the web UI menus.
  • Poller configuration generation + reload and restart method.
  • Schedule immediate checks (Central + Pollers) , acknowledgements, downtime, etc.
  • Move resources or reboot the active server and check again that everything is fine.