- large
- xlarge
- 2xlarge
- 4xlarge
- 8xlarge
- 10xlarge
- 12xlarge
- 16xlarge
- 24xlarge
- 32xlarge
Example of R5.
Start from db.r5.large
with 1 core, 2 vCPU and 16GB Memory.
Up to db.r5.24xlarge
48 Cores, 96 vCPU and 768GB Memory.
- T4g
- T3
- M7g
- M6i
- M6g
- R7g — up to 16xlarge
- R6i — up to 32xlarge
- R6id with NVMe-based SSD up to 7.6TB
- R6g — up to 16xlarge
- R5 — up to 24xlarge
R7g instances are powered by Arm-based AWS Graviton3 processors. DDR5 memory.
R6i instances are powered by 3rd-gen Intel Xeon Scalable processors (Ice Lake).
R6g instances are powered by Arm-based AWS Graviton2 processors.
General Purpose (SSD): gp2, gp3
Provisioned IOPS (SSD): low latency, I/O intensive
Magnetic: low cost per GB, backwards compatible
T3 is general purpose and M5 is better T3. R5 is better and memory optimized.
Either start with M5 or R5 for real-production use. R5 is better because the ceiling is higher in case your business succeeds.
You can find IOPS
in Databases > Configuration when your instance reached certain size.
Activity of RDS. Top SQL / most expensive queries. Adjustable to 5m, 1h, 5h, 24h, 1w, all. OS Metrics.
- Counter metrics
- Database load
- Top waits / Top SQL / Top hosts / Top users
Since PostgreSQL 10.
-
Databases > your-database-name > Monitoring (tab) > Monitoring (dropdown middle right) > Enhanced Monitoring
-
CloudWatch Logs > Log groups > RDSOSMetrics
CloudWatch metrics and alarms. Performance Insights actually exports metrics to CloudWatch.
- Regular snapshots for statistical data from PostgreSQL
Grant Superuser to a user/role:
GRANT rds_superuser TO user-or-role;