Building a High-Availability PostgreSQL Cluster at ARIN [Archived]
OUT OF DATE?
Here in the Vault, information is published in its final form and then not changed or updated. As a result, some content, specifically links to other pages and other references, may be out-of-date or no longer available.
By Devon Mizelle, Systems Administrator, ARIN
ARIN is a small organization with a big job. ARIN keeps track of IP addresses and Autonomous system numbers for both North America and parts of the Caribbean. This information is kept in a high-availability database that is core to ARIN’s services. This database is known as Jon Postel’s notebook, in tribute to Dr. Jon Postel who was one of the early luminaries of the Internet and who had the first job to keep track of names and numbers. Until the end of last year, this database resided in an Oracle cluster located in our base headquarters in Chantilly, VA. Over the past couple of years, ARIN has been looking to replace Oracle with the advent of robust Open Source database solution.
An organizational decision was made to migrate away from Oracle to PostgreSQL as it provided a feature set that we could use. One of the tougher requirements was replicating Oracle’s RAC system to provide us with high-availability for our backend. Through a long and intense period of research, implementation, and testing, we completed the migration to pgSQL late last year. So far, the new system has provided us great results.
This system is built upon a sturdy foundation made up of the new pgSQL replication functionality in the 9.0 release, the Corosync Cluster Engine, and the Pacemaker cluster resource management system. Using the Pacemaker system, we can monitor the health of each member of the database cluster and gracefully disable a misbehaving node in order to preserve the integrity of our services to the community. The Corosync engine provides the ‘back bone’ that the Pacemaker service uses to talk to each box and describe their health to one another. Essentially, we have been able to closely replicate the RAC system features that we used with Oracle using only open source software!
As a result of working with the Postgres community, we have been invited to speak at PGConf NYC 2014 about our new PostgresSQL cluster and how we’ve built it! We’re extremely excited about this opportunity and would like to extend an invitation to the rest of our community. If you are attending the conference, or are just in the area, please come out and say hello! You can find more information about the talk and the time it will be at on PGConf NYC’s site .
Any views, positions, statements or opinions of a guest blog post are those of the author alone and do not represent those of ARIN. ARIN does not guarantee the accuracy, completeness or validity of any claims or statements, nor shall ARIN be liable for any representations, omissions or errors contained in a guest blog post.
OUT OF DATE?
Here in the Vault, information is published in its final form and then not changed or updated. As a result, some content, specifically links to other pages and other references, may be out-of-date or no longer available.