An Ansible role for setup a percon XtraDB Cluster. Specifically, the responsibilities of this role are to:
- install packages
- secure connections
- bootstrap the cluster
No specific requirements
| Variable | Default | Comments (type) |
|---|---|---|
xtradb_bind_address |
- | The listening IP |
xtradb_bind_interface |
eth0 |
The interface used by the cluster |
xtradb_cluster_name |
foo |
A name for the cluster |
xtradb_configured |
xtradb_datadir/configured |
A cookie for idempotency |
xtradb_datadir |
/var/lib/mysql |
Directory of data |
xtradb_master_node |
groups[xtradb_nodes_group][0] |
The chosen node to be master |
xtradb_mysql_user |
mysql |
The user for run galera |
xtradb_nodes_group |
xtradb-cluster-nodes |
Node group where the cluster will be installed |
xtradb_root_password |
root |
Password for the root user |
xtradb_root_user |
root |
The root user |
xtradb_secured |
xtradb_datadir/secured |
A cookie for idempotency |
xtradb_service |
mysql |
Linux service name |
xtradb_sst_password |
sstpassword |
Password for the xtradb_sst_user |
xtradb_sst_user |
sstuser |
User used to the state snapshot transfer |
xtradb_swappiness |
0 |
"Swappiness" value. System default is 60. A value of 0 means that swapping out processes is avoided. |
xtradb_databases |
[] | List of names of the databases to be added |
xtradb_users |
[] | List of dicts specifying the users to be added. See below for details. |
xtradb_version |
57 |
Package version of XtraDB |
For more info on the values, read the MariaDB Server System Variables documentation.
| Variable | Default | Comments (type) |
|---|---|---|
xtradb_binlog_format |
ROW |
The binary logging format |
xtradb_character_set_server |
utf |
The character set |
xtradb_collation_server |
utf8_general_ci |
The collation |
xtradb_default_storage_engine |
InnoDB |
Setting the Storage Engine |
xtradb_innodb_autoinc_lock_mode |
2 |
There are three possible settings for the innodb_autoinc_lock_mode configuration parameter. The settings are 0, 1, or 2, for “traditional”, “consecutive”, or “interleaved” lock mode, respectively |
xtradb_innodb_buffer_pool_instances |
|
To enable multiple buffer pool instances, set the innodb_buffer_pool_instances configuration option to a value greater than 1 (8 is the default) up to 64 (the maximum). This option takes effect only when you set innodb_buffer_pool_size to a size of 1GB or more. The total size you specify is divided among all the buffer pools |
xtradb_innodb_buffer_pool_size |
|
A good value is 70%-80% of available memory. |
xtradb_innodb_file_format |
|
|
xtradb_innodb_file_format_check |
|
|
xtradb_innodb_file_per_table |
|
|
xtradb_innodb_flush_log_at_trx_commit |
|
When innodb_flush_log_at_trx_commit is set to 1 the log buffer is flushed on every transaction commit to the log file on disk and provides maximum data integrity but it also has performance impact. Setting it to 2 means log buffer is flushed to OS file cache on every transaction commit. The implication of 2 is optimal and improve performance if you are not concerning ACID and can lose transactions for last second or two in case of OS crashes. |
xtradb_innodb_log_buffer_size |
|
Innodb writes changed data record into lt’s log buffer, which kept in memory and it saves disk I/O for large transactions as it not need to write the log of changes to disk before transaction commit. 4 MB – 8 MB is good start unless you write a lot of huge blobs |
xtradb_innodb_log_file_size |
|
Default value has been changed in MySQL 5.6 to 50 MB from 5 MB (old default), but it’s still too small size for many workloads |
xtradb_innodb_file_per_table |
on |
innodb_file_per_table is ON by default from MySQL 5.6. This is usually recommended as it avoids having a huge shared tablespace and as it allows you to reclaim space when you drop or truncate a table. Separate tablespace also benefits for Xtrabackup partial backup scheme |
xtradb_innodb_strict_mode |
on |
|
xtradb_join_buffer_size |
|
|
xtradb_log_warnings |
|
|
xtradb_log_warnings |
|
|
xtradb_long_query_time |
|
|
xtradb_max_allowed_packet |
|
|
xtradb_max_connections |
4096 |
|
xtradb_max_heap_table_size |
|
|
xtradb_max_user_connections |
|
|
xtradb_pxc_strict_mode |
ENFORCING |
PXC Strict Mode is designed to avoid the use of experimental and unsupported features in Percona XtraDB Cluster |
xtradb_query_cache_size |
|
|
xtradb_read_buffer_size |
|
|
xtradb_read_rnd_buffer_size |
|
|
xtradb_skip_name_resolve |
1 |
Use IP addresses only. Set to 0 to resolve host names. |
xtradb_slow_query_log |
0 |
Set to 1 to enable the slow query log. |
xtradb_socket |
|
|
xtradb_sort_buffer_size |
|
|
xtradb_table_definition_cache |
|
|
xtradb_table_open_cache |
|
|
xtradb_table_open_cache_instances |
|
|
xtradb_tmp_table_size |
|
Databases are defined with a dict containing the fields name: (required), and init_script: (optional).
The init script is a SQL file that is executed when the database is created to initialise tables and populate it with values.
xtradb_databases:
- name: keystone
- name: mydb
init_script: files/init_mydb.sqlUsers are defined with a dict containing fields name:, password:, priv:, and, optionally, host:.
The password is in plain text and priv: specifies the privileges for this user as described in the Ansible documentation.
An example:
xtradb_users:
- name: keystone
password: KEYSTONE_DBPASS
priv: 'keystone.*:SUPER'
- name: cdelgehier
password: yolo
priv: 'mydb.*:ALL'
host: '192.168.1.%'No dependencies.
- hosts: db
gather_facts: true
become: true
roles:
- role: ansible-role-XtraDB-Cluster
xtradb_cluster_name: "prod-customer"
xtradb_sst_user: sstuser
xtradb_sst_password: s3cr3t
xtradb_root_password: yolo
xtradb_nodes_group: "db"
xtradb_bind_interface: eth0- hosts: db
gather_facts: true
become: true
roles:
- role: ansible-role-XtraDB-Cluster
xtradb_cluster_name: "prod-customer"
xtradb_sst_user: sstuser
xtradb_sst_password: s3cr3t
xtradb_root_password: yolo
xtradb_bind_address: "{{ ansible_default_ipv4.address }}"
xtradb_wsrep_cluster_address: "gcomm://{{ groups['db'] | map('extract', hostvars, ['ansible_default_ipv4', 'address']) | join(',') }}"
xtradb_master_node: "{{ hostvars[ groups['db'][0] ].ansible_default_ipv4.address }}"
- hosts: db
gather_facts: true
become: true
roles:
- role: ansible-role-XtraDB-Cluster
xtradb_cluster_name: "prod-customer"
xtradb_sst_password: s3cr3t
xtradb_root_password: yolo
xtradb_nodes_group: "db"
xtradb_bind_interface: eth0
xtradb_databases:
- name: keystone
xtradb_users:
- name: keystone
password: PASSWD
priv: 'keystone.*:GRANT,ALL'
xtradb_innodb_buffer_pool_instances: 8
xtradb_innodb_buffer_pool_size: "384M"
xtradb_innodb_file_format: "Barracuda"
xtradb_innodb_file_format_check: "1"
xtradb_innodb_file_per_table: "on"
xtradb_innodb_flush_log_at_trx_commit: "1"
xtradb_innodb_log_buffer_size: "16M"
xtradb_innodb_log_file_size: "50M"
xtradb_innodb_strict_mode: "on"
xtradb_join_buffer_size: "128K"
xtradb_log_warnings: "1"
xtradb_long_query_time: "10"
xtradb_max_allowed_packet: "16M"
xtradb_max_connections: "505"
xtradb_max_heap_table_size: "16M"
xtradb_max_user_connections: "500"
xtradb_query_cache_size: "0" # disable
xtradb_read_buffer_size: "128K"
xtradb_read_rnd_buffer_size: "256k"
xtradb_skip_name_resolve: "1"
xtradb_slow_query_log: "1"
xtradb_sort_buffer_size: "2M"
xtradb_table_definition_cache: "1400"
xtradb_table_open_cache: "2000"
xtradb_table_open_cache_instances: "8"
xtradb_tmp_table_size: "16M"
[db]
node1 ansible_host=192.168.1.173
node2 ansible_host=192.168.1.156
node3 ansible_host=192.168.1.154There are two types of test environments available. One powered by Vagrant, another by Docker. The latter is suitable for running automated tests on Travis-CI. Test code is kept in separate orphan branches. For details of how to set up these test environments on your own machine, see the README files in the respective branches:
- Docker: docker-tests
To remove alltraces and start a new install
ansible db -m shell -a 'rm -rf /var/lib/mysql /var/log/mysqld.log /etc/percona-xtradb-cluster.conf.d ; yum remove Percona* -y'
Issues, feature requests, ideas are appreciated and can be posted in the Issues section.
Pull requests are also very welcome. The best way to submit a PR is by first creating a fork of this Github project, then creating a topic branch for the suggested change and pushing that branch to your own fork. Github can then easily create a PR based on that branch.
2-clause BSD license, see LICENSE.md
- Cedric DELGEHIER (maintainer)