AWS · Software · YoctoIT tech page

RDS

Managed relational databases: MySQL, PostgreSQL, SQL Server and Oracle without basic administration — and Aurora, AWS's cloud-native engine.

FOCUS · THE DBA SLEEPSAutomatic patching, backups and failover: the database that doesn't call at night
Request the Free Assessment Learn more
YoctoIT material for clients and partners · AWS e i nomi dei servizi sono marchi di Amazon.com, Inc.
01 · What it is

Amazon RDS & Aurora, made clear.

RDS removes the DBA's repetitive work: provisioning, patching, point-in-time backups and automatic multi-AZ failover for the main engines. Aurora redesigns MySQL and PostgreSQL on AWS's distributed storage: more throughput, replicas in seconds and Serverless v2.

6 engines
MySQL, PostgreSQL, MariaDB, SQL Server, Oracle, Db2
PITR
point-in-time recovery: the database rewound to the second
15 replicas
Aurora read replicas at minimal latency
Icona ufficiale AWS — Amazon RDS & Aurora
OFFICIAL AWS ICON · RDS
Console AWS
REAL AWS CONSOLE · RDS PERFORMANCE INSIGHTS · SOURCE: AWS BLOG
02 · How to use it well

The things that make the difference.

High availability

Applicationsingle endpoint, transparent failover
Primary AZ-a
Standby AZ-b
Read replicas
native multi-AZ
Distributed Aurora storage6 copies across 3 AZs, automatic repair
Backups on S3 + PITRcontinuous protection, no windows
The database that survives the zone

Multi-AZ

The synchronous standby with failover in ~1-2 minutes: HA without a cluster to build.

Aurora Serverless v2

Capacity scaling in fine steps: the database that follows the load, even to zero.

Blue/Green deployment

Version upgrades rehearsed on the clone and swapped in seconds: the major upgrade without terror.

Performance Insights

Visual tuning: heavy queries seen immediately, no AWR to decipher.

03 · In depth

Multi-AZ, Aurora and the versions that expire

RDS manages MySQL, PostgreSQL, MariaDB, SQL Server and Oracle with Multi-AZ (automatic failover ~60-120s; with Multi-AZ cluster, two readable replicas and faster failover) and PITR up to 35 days. Aurora separates storage and compute: 6 copies across 3 AZs, replicas at minimal lag, Serverless v2 scaling in 0.5 ACU steps and Global Database for cross-region DR with ~1s RPO. Engine versions have an end of life: the calendar is part of the service.

  • Multi-AZ cluster — two readable standbys and ~35s failover: RDS's modern HA
  • Aurora storage — 6 self-healing copies across 3 AZs: the volume growing by itself up to 128TB
  • Serverless v2 — fine scaling at 0.5 ACU: the database that breathes with the load
  • Global Database — cross-region replication with ~1s RPO: managed geographic DR
  • Performance Insights — load by query and waits: data-driven tuning
  • Blue/Green deploy — the version upgrade with a minutes-long switch and rollback ready
04 · Numbers and lifecycle

The numbers that matter.

35 gg
the maximum point-in-time recovery
~1 s
the typical Aurora Global Database RPO
128 TB
the maximum Aurora volume per cluster
EOL
every engine major has a date: we govern it by calendar
The managed database leaves you schema and versions: parameters, failover tests and blue/green upgrades are in our perimeter.
05 · Use cases

Where it really pays off.

The ERP in the cloud

The application database on RDS: the migration that frees the team from the iron.

New applications

Aurora for cloud-born projects: performance and resilience from day one.

Instance consolidation

Dozens of scattered databases brought back to standard: single policies, visible costs.

The database remains the heart: we watch it with Yocto Vision and keep it at the right levels — RDS does the manual labor, we do the strategy.