Permalink
Switch branches/tags
Find file Copy path
Fetching contributors…
Cannot retrieve contributors at this time
115 lines (86 sloc) 5.16 KB

About Galera Cluster for MySQL

Galera Cluster for MySQL is a synchronous, multi-master cluster for the MySQL database. You can read and write to any node, and you can lose any one node without interruption in operations, and without the need to handle complicated failover procedures.

/images/galera_multi_master.png

Galera Overview

Galera Replication is a general purpose replication library for any transactional system. It can be used to create a synchronous multi-master replication solution, such as to achieve high availability and scale-out.

The first and most popular Galera based solution is Galera Cluster for MySQL. On a high level, this is standard MySQL, and standard InnoDB storage engine, using the Galera replication library as a replication plugin. To be precise, we do extend the MySQL replication plugin API to provide all the information and hooks needed for truly multi-master, synchronous replication. We call this API the Write Set REPlication API. [1] To learn more about the internals of Galera, WSREP API and how they interact with MySQL, you should read :doc:`understanding/internals`.

To use Galera Cluster for MySQL, you need to download a version of MySQL that includes the WSREP patches. Galera enabled versions are available for all major MySQL forks: Oracle MySQL, Percona Server and MariaDB. The chapter :doc:`installing` contains links to download each of these.

Galera implements so called certification based replication. This concept was introduced in, and Galera was inspired by, the academic research of Fernando Pedone. The basic idea is that the transaction to be replicated - the write set - not only contains the database rows to replicate, it also includes information about all the locks that were held by the database (ie. InnoDB) during the transaction. Each node then certifies the replicated write set against other write sets in the applier queue, and if there are no conflicting locks, we know that the write set can be applied. At this point the transaction is considered committed, after which each node continues to apply it to the InnoDB tablespace.

This approach is also called virtually synchronous replication since it is logically synchronous, but actual writing (and committing) to the InnoDB tablespace happens independently (and thus, strictly speaking, asynchronously) on each node.

Galera benefits

Galera is a significant step in terms of improving MySQL high availability. The MySQL ecosystem has offered users various ways to achieve high availability, but typically all alternatives have provided only some of the following features, but not all, thus making the choice of a high availability solution an exercise in tradeoffs.

Galera provides all of the following features:

  • True multi-master. Read and write to any node at any time.
  • Synchronous replication. No data is lost at node crash or failover.
  • Tightly coupled. All nodes hold the same state. No diverged data between nodes allowed.
  • Multi-threaded slave, for any workload. No slave lag or performance overhead.
  • No master-slave failover operations or use of VIP necessary.
  • Hot standby. No downtime during failover (since there is no failover).
  • Automatic node provisioning. No need to manually back up the database and copy it to the new node.
  • Supports InnoDB storage engine. Compatible with MySQL installed base.

The result is a high availability solution that is both robust in terms of data integrity and high performance with instant failovers. The following diagram illustrates how different common database clustering solutions compare in these dimensions:

/images/ha_alternatives.png

The other main benefit of Galera is good support for typical cloud usage. Automatic node provisioning makes elastic scale-out and scale-in operations painless. Galera has proven to perform very well in the cloud, such as when using multiple small node instances, across multiple data centers (e.g. AWS zones) or even over Wider Area Network.

About Codership Oy

Galera was created by Codership Oy, a company based in Helsinki, Finland.

Codership was founded and development on Galera was started in 2007 by the three founders Seppo Jaakola, Teemu Ollakka and Alex Yurchenko. Development was funded by doing consulting work, and later by startup grants and loans from Finnish government agency Tekes. Before starting work on Galera, the founders had already worked on 3 previous clustering solutions for MySQL, the last of which, Tungsten, is also in widespread use today.

Galera 1.0 was eventually released in October 2011. At this point there were already several customers running Galera Cluster for MySQL in production on the beta versions 0.7 and 0.8.

The word "Codership" is based on the word "coder" / "coders". Think of it as similar to the word "friendship".

Footnotes

[1]To use the Galera replication library in some other database or transactional application than MySQL, you would similarly implement this WSREP API.