Archive

The Dulin Report

Browsable archive from the WordPress export.

2015

On Managing Stress, Multitasking and Other New Year's Resolutions Jan 1, 2015 Configuring Master-Slave Replication With PostgreSQL Jan 31, 2015 Trying to Replace Cassandra with DynamoDB ? Not so fast Feb 2, 2015 On apprenticeship Feb 13, 2015 Where AWS Elastic BeanStalk Could be Better Mar 3, 2015 Finding Unused Elastic Load Balancers Mar 24, 2015 Do not apply data science methods without understanding them Mar 25, 2015 Microsoft and Apple Have Everything to Lose if Chromebooks Succeed Mar 31, 2015 Two developers choose to take a class Apr 1, 2015 What can Evernote Teach Us About Enterprise App Architecture Apr 2, 2015 Exploration of the Software Engineering as a Profession Apr 8, 2015 Ordered Sets and Logs in Cassandra vs SQL Apr 8, 2015 Building a Supercomputer in AWS: Is it even worth it ? Apr 13, 2015 Apple is (or was) the Biggest User of Apache Cassandra Apr 23, 2015 My Brief Affair With Android Apr 25, 2015 Why I am not Getting an Apple Watch For Now: Or Ever Apr 26, 2015 The Clarkson School Class of 2015 Commencement May 5, 2015 The Clarkson School Class of 2015 Commencement speech May 5, 2015 We Need a Cloud Version of Cassandra May 7, 2015 Guaranteeing Delivery of Messages with AWS SQS May 9, 2015 Smart IT Departments Own Their Business API and Take Ownership of Data Governance May 13, 2015 Big Data is not all about Hadoop May 30, 2015 The longer the chain of responsibility the less likely there is anyone in the hierarchy who can actually accept it Jun 7, 2015 Your IT Department's Kodak Moment Jun 17, 2015 Attracting STEM Graduates to Traditional Enterprise IT Jul 4, 2015 Book Review: "Shop Class As Soulcraft" By Matthew B. Crawford Jul 5, 2015 The Three Myths About JavaScript Simplicity Jul 10, 2015 Social Media Detox Jul 11, 2015 Big Data Should Be Used To Make Ads More Relevant Jul 29, 2015 On Maintaining Personal Brand as a Software Engineer Aug 2, 2015 Ten Questions to Consider Before Choosing Cassandra Aug 8, 2015 What Every College Computer Science Freshman Should Know Aug 14, 2015 We Live in a Mobile Device Notification Hell Aug 22, 2015 Top Ten Differences Between ActiveMQ and Amazon SQS Sep 5, 2015 Setting Up Cross-Region Replication of AWS RDS for PostgreSQL Sep 12, 2015 I Stand With Ahmed Sep 19, 2015 Banking Technology is in Dire Need of Standartization and Openness Sep 28, 2015 IT departments must transform in the face of the cloud revolution Nov 9, 2015 Operations costs are the Achille's heel of NoSQL Nov 23, 2015 Our civilization has a single point of failure Dec 16, 2015

Setting Up Cross-Region Replication of AWS RDS for PostgreSQL

September 12, 2015

As of today AWS RDS for PostgreSQL1 does not offer cross-region replication. Short of switching to one of the RDS offerings that do support it, there is a few options to consider.

1. Custom Configured EC2 Instances with Master-Slave Replication


[caption id="attachment_222" align="aligncenter" width="300"]Master/slave light rail Master/slave light rail[/caption]

This setup sacrifices the benefits of AWS RDS service in exchange for greater control over replication settings. In this setup, one region hosts a master PostgreSQL host, and another region hosts a slave which can also act as a read-replica2.

Advantages


Greater control over replication settings.

Disadvantages



  • Give up all the advantages of running in AWS RDS environment.

  • Writes can only be performed in the master region.


2. Software-defined Two-phase Commit


[caption id="attachment_264" align="aligncenter" width="300"]Commitment Photo credit: Ed Schipul Commitment
Photo credit: Ed Schipul[/caption]

In this setup there are two independent AWS RDS instances. The application, however, utilizes a two-phase commit protocol3 to guarantee that all writes make it into both databases in a transactional fashion.

Advantages



  • Simple configuration

  • Does not sacrifice any of the AWS RDS advantages


Disadvantages



  • Responsibility for ensuring that writes make it into all regions fall onto the application itself.

  • Increased application code complexity.

  • Write performance is sacrificed since all regional databases must participate synchronously.


3. Asynchronous Writers


[caption id="attachment_266" align="aligncenter" width="300"]Fanning Out Photo credit: Tim Haynes Fanning Out
Photo credit: Tim Haynes[/caption]

In this approach each region hosts an asynchronous writer that listens on an SQS queue4. All writes are published on the SNS topic that is configured with all regional writer queues as subscriptions5. When the application running in any of the regions wants to write into the database it publishes a message on this SNS topic which then fans it out to all of the regional SQS queues.

Advantages



  • Simple configuration

  • Does not sacrifice any of the AWS RDS advantages

  • Does not sacrifice write performance


Disadvantages



  • Subject to software bugs

  • Subject to SNS and SQS bugs and outages

  • No guarantee of consistency

  • Requires a mechanism for periodically reconciling differences between regions







  1. AWS RDS for PostgreSQL ↩

  2. Configuring Master/Slave Replication with PostgreSQL ↩

  3. Distributed Transactions and the Transaction Manager ↩

  4. Amazon Simple Queue Service Documentation ↩

  5. SQS Queues and SNS Notifications – Now Best Friends ↩


Last updated January 20, 2026. Return to home.