Like all Compose databases, PostgreSQL Deployments allow developers to focus on what they do best (developing useful applications) and avoid the guesswork of capacity planning. These deployments are deployed on fast SSDs, fully replicated, and have access to huge amounts of RAM. As an application’s data size increases, the deployment resources scale fluidly to maintain consistent performance, and are billed based on actual usage.
PostgreSQL uses RAM and I/O efficiently. Because of this, we scale RAM at a 1-to-10 ratio with datasize. As your data size increases by 1GB, we include 100MB of RAM.
- Uses the latest production PostgreSQL 9.4 and 9.5 release
- Automatically scaling RAM, CPU, and I/O as your PostgreSQL data grows.
- Guaranteed resources per deployment.
- 2 data-nodes clusters replicated using WAL streaming
- Automatic failover process begins after leader node is down for 60 seconds
- Running on extremely fast SSD (solid state drives)
- Access to the standard Contrib add-ons
- No limit on database creation or schema create - you have CREATEDB and CREATEROLE access
- No-cost backups, no matter how big your deployment grows.
- Enabled SSL connections
- Start with 1GB for $17.50 - as you grow each additional gigabyte of data costs $12.
- Ubuntu: Run
sudo apt-get install postgresql-clientto install the psql command.
- Mac OS X:
- Other platforms: Consult the PostgreSQL download page for details of the various install options.
See Compose Datacenter Availability for current location availability.
PostgreSQL Deployments include the following automatic daily backup pattern:
- Daily backups for 7 days
- Weekly backups for 4 weeks
- Monthly backups for 3 months
Backups include all PostgreSQL data files, and are available for download. Backup downloads include a README file. See this file for instructions on restoring the data.
All PostgreSQL Deployments are high-availability clusters with 2 data nodes. Data nodes are powered by fully redundant architecture from the datacenter to the server. PostgreSQL clusters consist of a leader and a follower member on separate machines in separate zones. In the event of a leader member failure, we've configured automatic promotion of the current slave within 60 seconds of the leader becoming unresponsive. Failover will only occur if the follower is within 1 MB of the leader, else the deployment will wait for the leader to come online. Using the Deployment topology on your PostgreSQL deployment to see if your follower is healthy enough to become Primary.
Each PostgreSQL deployment has a proxy, which provides a single member connection string so there's no worrying with which member is currently leader. Failover will happen seamlessly without requiring you to change your application's connection configuration.