Atlassian has migrated its four million Postgres databases backing customer Jira implementations to Amazon Web Services' Aurora. The company operates one database per Jira customer, utilizing about 3,000 PostgreSQL servers across 13 AWS regions. Transitioning to Aurora PostgreSQL targets reduced AWS instance sizes, enhanced uptime from 99.95% to 99.99%, and better autoscaling features. The instance size changed from m5.4xlarge on RDS to r6.2xlarge on Aurora, maintaining memory while reducing CPU count. The complex migration met ambitious cost-saving goals and improved system reliability and performance.
Atlassian migrated four million Postgres databases supporting its Jira implementations to AWS Aurora to enhance cost, reliability, and performance metrics.
Rubis revealed that Atlassian operationalized one database for each Jira customer across about 3,000 PostgreSQL servers in 13 AWS regions.
The move from AWS RDS to Aurora PostgreSQL will enable Atlassian to cut AWS instance sizes by half and improve uptime from 99.95% to 99.99%.
Despite the migration being technically challenging, Atlassian successfully achieved its ambitious cost-saving targets while improving reliability and performance.
Collection
[
|
...
]