Aurora PostgreSQL

Amazon Aurora PostgreSQL is a PostgreSQL-compatible database with Amazon’s own distributed storage layer underneath. The SQL surface, the wire protocol and the system catalog are PostgreSQL’s, so AddisDB connects with its Postgres driver and everything downstream behaves as it does on stock Postgres.

Aurora DSQL is an entirely different product — serverless, distributed, with a deliberately reduced catalog — and it has its own picker entry. If your endpoint ends in dsql.amazonaws.com, use Aurora DSQL rather than this one.

Who it is for

Aurora suits workloads that outgrow a single RDS instance on read throughput or failover time. Storage is replicated six ways across three availability zones and grows on its own, and a replica promotes in seconds rather than minutes.

The detail that matters when you connect is that a cluster has two endpoints. The writer endpoint always points at the current primary; the reader endpoint load-balances across replicas and is read-only. Pointing AddisDB at the reader and then wondering why an INSERT fails is the classic first mistake — the error comes from Aurora, not from the app.

Set up the cluster

1. In the RDS console open your Aurora cluster and copy the writer endpoint — it contains cluster- in the hostname.
2. Confirm the port, 5432 by default.
3. Make sure you are inside the VPC or that the cluster is publicly accessible, and that the security group allows your address on 5432.
4. Note the master username and the initial database name.

# The writer endpoint is the one that accepts writes
psql -h my-cluster.cluster-abc123.us-east-1.rds.amazonaws.com -p 5432 -U postgres -c 'SELECT pg_is_in_recovery();'

That query answers the endpoint question directly: false means you are on the writer, true means you reached a reader.

Connect from AddisDB

1. New Connection → Aurora PostgreSQL. Port prefills to 5432, database and username to postgres.
2. Paste the writer endpoint into Host unless you deliberately want a read-only session.
3. Set SSL mode to require, or verify-full with the AWS regional CA bundle selected as the CA certificate.
4. If you are connecting to a reader endpoint on purpose, mark the connection read-only in AddisDB as well so the app stops writes before Aurora has to.
5. Test, then Save.

What AddisDB gives you

The full PostgreSQL toolset: schema tree, in-grid editing, structured create and alter table, database users, mock data, and the query plan viewer with the complete EXPLAIN option list including buffers and per-node timings.

The Live Monitor is real — Aurora exposes a genuine pg_stat_activity, so live sessions, table sizes and cancelling or terminating a backend all work. Test clones work through CREATE DATABASE … TEMPLATE.

For statement statistics, enable pg_stat_statements in the cluster parameter group, reboot once, then CREATE EXTENSION pg_stat_statements. Aurora reports it per instance, so profile against the writer if you want the write traffic included.