galera cluster

Add to new model

Galera Cluster provides an active/active MySQL
compatible alternative implemented using the Galera
synchronous replication library.

This charm is based on the Percona Cluster charm by
James Page james.page@ubuntu.com

Tags

Overview

Galera Cluster is a high availability and high scalability solution for
MySQL clustering. Galera Cluster integrates MySQL with the
Galera library of MySQL high availability solutions in a single product package
which enables you to create a cost-effective MySQL cluster.

This charm deploys Galera Cluster onto Ubuntu.

Usage

WARNING: Its critical that you follow the bootstrap process detailed in this
document in order to end up with a running Active/Active Galera Cluster.

Configuration

The following configuration parameters are required:

  • root-password - Due to the multi-master nature of Galera Cluster, the root-password
    needs to be specified in advance so that all nodes have the same root password.

  • sst-password - This password is used as a shared secret between nodes when performing
    whole-database transfers between nodes during node deployment.

Proxy Configuration

If you are deploying this charm on MAAS or in an environment without direct
access to the internet, you will need to allow access to releases.galeracluster.com
as the charm installs packages directly from the Galera repositories. If you
are using squid-deb-proxy, follow the steps below:

echo "releases.galeracluster.com" | sudo tee /etc/squid-deb-proxy/mirror-dstdomain.acl.d/40-galera
sudo service squid-deb-proxy restart

Deployment

The first service unit deployed acts as the seed node for the rest of the
cluster; in order for the cluster to function correctly, the same MySQL passwords
must be used across all nodes:

cat > galera.yaml << EOF
galera-cluster:
    root-password: my-root-password
    sst-password: my-sst-password
EOF

Once you have created this file, you can deploy the first seed unit:

juju deploy --config galera.yaml cs:~codership/galera-cluster

Once this node is full operational, you can add extra units one at a time to the
deployment:

juju add-unit galera-cluster

A minimum cluster size of three units is recommended.

In order to access the cluster, you can use the hacluster charm in multicast-capable
deployments to provide a single IP address:

juju set galera-cluster vip=10.0.3.200
juju deploy hacluster
juju add-relation hacluster galera-cluster

Clients can then access using the vip provided. This vip will be passed to
related services:

juju add-relation keystone galera-cluster

Limitations

Note that Galera Cluster is not a 'scale-out' MySQL solution; reads
and writes are channeled through a single service unit and synchronously
replicated to other nodes in the cluster; reads/writes are as slow as the
slowest node you have in your deployment.

Configuration

vip_iface
(string) 
                            Network interface on which to place the Virtual IP
eth0
root-password
(string) 
                            Root password for MySQL access; must be configured pre-deployment for Active-Active clusters.
source
(string) 
                            Package install location for Galera Cluster (defaults to distro for >= 14.04)
innodb-file-per-table
(boolean) 
                            Turns on innodb_file_per_table option, which will make MySQL put each
InnoDB table into separate .idb file. Existing InnoDB tables will remain
in ibdata1 file - full dump/import is needed to get rid of large
ibdata1 file
True
vip_cidr
(int) 
                            Netmask that will be used for the Virtual IP
24
sst-password
(string) 
                            Re-sync account password for new cluster nodes; must be configured pre-deployment for Active-Active clusters.
access-network
(string) 
                            The IP address and netmask of the 'access' network (e.g., 192.168.0.0/24)
.
This network will be used for access to database services.
vip
(string) 
                            Virtual IP to use to front Galera Cluster in active/active HA configuration
ha-mcastport
(int) 
                            Default multicast port number that will be used to communicate between HA Cluster nodes.
5490
max-connections
(int) 
                            Maximum connections to allow. -1 means use the server's compiled in default.
-1
prefer-ipv6
(boolean) 
                            If True enables IPv6 support. The charm will expect network interfaces
to be configured with an IPv6 address. If set to False (default) IPv4
is expected.
.
NOTE: these charms do not currently support IPv6 privacy extension. In
order for this charm to function correctly, the privacy extension must be
disabled and a non-temporary address must be configured/available on
your network interface.
dataset-size
(string) 
                            How much data do you want to keep in memory in the DB. This will be used to tune settings in the database server appropriately. Suffix this value with 'K','M','G', or 'T' to get the relevant kilo/mega/etc. bytes. If suffixed with %, one will get that percentage of RAM devoted to dataset.
80%
ha-bindiface
(string) 
                            Default network interface on which HA cluster will bind to communication with the other members of the HA Cluster.
eth0