3 Node Cluster

Overview

In this example, we have prepared 3 machines with Red Hat 9 operating system in the same network:

• 10.1.1.2 vprotect1.local

• 10.1.1.3 vprotect2.local

• 10.1.1.4 vprotect3.local

We will use IP 10.1.1.5/24 (storware.local) for the floating IP of our cluster.

On each machine, we are going to install Storware Server and Node. Cluster will work in Active-Passive mode, which means Storware Services will be active only on a single host of the cluster.

Storware server installation

Run those steps on all machines under the pacemaker cluster.

• Please follow the Storware server installation, assuming the Red Hat Enterprise Linux 9 OS: Server

• During installation, do not execute the command systemctl start vprotect-server

Steps:

• Create a repository file for Storware Backup and Recovery-

• Create a repository file for MariaDB

• Server installation (for Red Hat 9)

• Server configuration - do not execute systemctl start vprotect-server

Next, add firewall rules:

1. Add redirection to allow the local node to communicate with the server on the cluster IP

firewall-cmd --permanent --direct --add-rule ipv4 nat OUTPUT 0 -p tcp -o lo --dport 443 -j REDIRECT --to-ports 8181
firewall-cmd --complete-reload

1. Open firewall for MariaDB replication:

firewall-cmd --add-port=3306/tcp --permanent
firewall-cmd --complete-reload

Configuration custom SSL certificate

All steps run as the root user. All steps execute on the first node of the cluster. In our example, we are going to create a certificate for the domain storware.local

Follow the steps from enabling HTTPS connectivity for nodes

Storware node installation

• Please follow the Storware node installation, assuming the Red Hat Enterprise Linux 9 OS: Node

• During installation do not execute commands: systemctl start vprotect-node reboot vc node inst register

Steps:

• Node installation (For Red Hat 9)

• Node configuration - Only point 3 "Configure the operating system."

Backup destination configuration

For a multi-node/cluster environment for backup destination, we suggest using NFS, Object Storage, and Deduplication appliances. In this example, we use NFS. Execute on all Storware node machines.

1. Add an entry in /etc/fstab for automount NFS

10.1.1.1:/vprotect /vprotect_data nfs defaults 0 0

1. Create directories for the mount NFS share:

mkdir /vprotect_data

1. Mount the NFS share

mount -a

1. Create subdirectories for backup destinations (run only on a single node)

mkdir /vprotect_data/backup
mkdir /vprotect_data/backup/synthetic
mkdir /vprotect_data/backup/filesystem
mkdir /vprotect_data/backup/dbbackup

1. Add privileges for newly created shares

chown vprotect:vprotect -R /vprotect_data

Cluster Configuration

The cluster is controlled by a pacemaker.

Prepare the operating system

All steps run as the root user. Run those steps on all machines in the pacemaker cluster:

1. Stop all services controlled by the cluster, and disable autostart.

systemctl stop vprotect-node
systemctl stop vprotect-server
systemctl disable vprotect-node
systemctl disable vprotect-server

1. Configure DNS server, or the hosts file on each machine vi /etc/hosts

10.1.1.2 vprotect1.local
10.1.1.3 vprotect2.local
10.1.1.4 vprotect3.local
10.1.1.5 storware.local

Set MariaDB replication

All steps run as the root user. Run on all cluster nodes:

1. Create MariaDB user replication with password vPr0tect for replication:

CREATE USER replicator@'10.1.1.%' IDENTIFIED BY 'vPr0tect';
GRANT SELECT,REPLICATION SLAVE,REPLICATION CLIENT ON *.* to replicator@'10.1.1.%' IDENTIFIED BY 'vPr0tect';

1. Add changes to /etc/my.cnf.d/server.cnf in the mysqld section:

[mysqld]
lower_case_table_names=1
log-bin=mysql-bin
relay-log=relay-bin
log-slave-updates
max_allowed_packet=500M
log_bin_trust_function_creators=1

1. Add changes to /etc/my.cnf.d/server.cnf in the mysqld section: On vprotect1.local:

server-id=10

On vprotect2.local:

server-id=20

On vprotect3.local:

server-id=30

1. Restart MariaDB service:

systemctl restart mariadb

1. On each host, show output from the command:

SHOW MASTER STATUS;

Output from vprotect3.local:

MariaDB [(none)]> SHOW MASTER STATUS;
+------------------+----------+--------------+------------------+
| File             | Position | Binlog_Do_DB | Binlog_Ignore_DB |
+------------------+----------+--------------+------------------+
| mysql-bin.000001 |      328 |              |                  |
+------------------+----------+--------------+------------------+
1 row in set (0.000 sec)

Output from vprotect1.local:

MariaDB [(none)]> SHOW MASTER STATUS;
+------------------+----------+--------------+------------------+
| File             | Position | Binlog_Do_DB | Binlog_Ignore_DB |
+------------------+----------+--------------+------------------+
| mysql-bin.000001 |      328 |              |                  |
+------------------+----------+--------------+------------------+
1 row in set (0.000 sec)

Output from vprotect2.local:

MariaDB [(none)]> SHOW MASTER STATUS;
+------------------+----------+--------------+------------------+
| File             | Position | Binlog_Do_DB | Binlog_Ignore_DB |
+------------------+----------+--------------+------------------+
| mysql-bin.000001 |      328 |              |                  |
+------------------+----------+--------------+------------------+
1 row in set (0.000 sec)

1. Set replication on each MariaDB server: Execute on vprotect1.local:

CHANGE MASTER TO
MASTER_HOST='10.1.1.4',
MASTER_PORT=3306,
MASTER_USER='replicator',
MASTER_PASSWORD='vPr0tect',
MASTER_LOG_FILE='mysql-bin.000001',
MASTER_LOG_POS=328;

Execute on vprotect2.local:

CHANGE MASTER TO
MASTER_HOST='10.1.1.2',
MASTER_PORT=3306,
MASTER_USER='replicator',
MASTER_PASSWORD='vPr0tect',
MASTER_LOG_FILE='mysql-bin.000001',
MASTER_LOG_POS=328;

Execute on vprotect3.local:

CHANGE MASTER TO
MASTER_HOST='10.1.1.3',
MASTER_PORT=3306,
MASTER_USER='replicator',
MASTER_PASSWORD='vPr0tect',
MASTER_LOG_FILE='mysql-bin.000001',
MASTER_LOG_POS=328;

1. Start replication MariaDB: Execute on vprotect1.local:

START SLAVE;

Show output from command:

SHOW SLAVE STATUS\G

Wait until you see in the output:

Slave_IO_Running: Yes
Slave_SQL_Running: Yes

Repeat the last step on host vprotect2.local and vprotect3.local.

Set the same passwords for the vprotect user in MariaDB

Execute only on first node of cluster.

1. Copy password from file /opt/vprotect/server/quarkus.properties

quarkus.datasource.password=SECRETPASSWORD

1. Log in to MariaDB

mysql -u root -p

1. Set password for vprotect user:

SET PASSWORD FOR 'vprotect'@'localhost' = PASSWORD('SECRETPASSWORD');
quit;

1. Copy configuration files from vprotect1.local to other cluster hosts

cd /opt/vprotect/server/
keystore.jks
log4j2-server.xml
quarkus.properties
vprotect.env
vprotect-keystore.jks
license.key

1. Add permissions for copied files (execute on each host)

chown vprotect:vprotect -R /opt/vprotect/

Configure pacemaker

All steps run as the root user.

Run on every node in the cluster

1. Install pacemaker packages

dnf --enablerepo=highavailability -y install pacemaker pcs

1. Create SSH keys, and add them to other hosts to know. Create SSH key on each host:

ssh-keygen

To add a certificate to know on another host: run on host1:

ssh-copy-id 10.1.1.3
ssh-copy-id 10.1.1.4

run on host2:

ssh-copy-id 10.1.1.2
ssh-copy-id 10.1.1.4

run on host3:

ssh-copy-id 10.1.1.2
ssh-copy-id 10.1.1.3

1. Open ports on the firewall

firewall-cmd --permanent --add-service=high-availability
firewall-cmd --reload

1. Start pcsd service

systemctl start pcsd.service
systemctl enable pcsd.service

1. Set the same password for the user hacluster

passwd hacluster

Run only on the first node of the cluster

1. Authenticate nodes of the cluster (with hacluster user)

pcs host auth vprotect1.local vprotect2.local vprotect3.local

1. Create cluster

pcs cluster setup vp vprotect1.local vprotect2.local vprotect3.local

1. Run cluster

pcs cluster start --all

1. Power off stonith

pcs property set stonith-enabled=false

1. Create a floating IP in the cluster

pcs resource create vp-vip1 IPaddr2 ip=10.1.1.5 cidr_netmask=24 --group vpgrp

1. Add vprotect-server to the cluster

pcs resource create "vp-vprotect-server.service" systemd:vprotect-server.service op start on-fail="stop" timeout="300s" op stop timeout="300s" op monitor timeout="300s" --group vpgrp

1. Register Storware node on server (run on all hosts) 7.1. Add the certificate to the trusted

/opt/vprotect/node/scripts/node_add_ssl_cert.sh 10.1.1.5 443

7.2. Register node on server

vc node inst register --name=${HOSTNAME%%.*} --login=admin --password=vPr0tect --apiurl=https://storware.local:443/api

1. Add vprotect-node to the cluster

pcs resource create "vp-vprotect-node.service" systemd:vprotect-node.service op start on-fail="stop" timeout="300s" op stop timeout="300s" op monitor timeout="300s" --group vpgrp

Useful commands to control the cluster:

For update, or service Storware unmanage services from cluster:

pcs resource unmanage vpgrp

Back to manage:

pcs resource manage vpgrp

Show status of cluster:

pcs status

Stop cluster node:

pcs cluster stop vprotect1.local

Stop all nodes of the cluster:

pcs cluster stop --all

Start all nodes of the cluster:

pcs cluster start --all

Clear old errors in cluster:

pcs resource cleanup