- Newest
- Most votes
- Most comments
Hello Krishnakumar,
For upgrading your Aurora MySQL cluster from db.r6g.large to db.r6g.12xlarge, this is an instance class change (vertical scaling) rather than an engine version upgrade, so the process is more straightforward. However, here are some important considerations:
Key Checks Before Upgrading:
-
Parameter Groups: If you're using a custom parameter group, verify that all parameters are appropriate for the larger instance class. Some parameters may need adjustment based on the increased resources (memory, CPU). If you're using the default parameter group, it will automatically apply to the new instance class.
-
Memory-Related Considerations: Monitor your current instance's FreeableMemory metric in CloudWatch. While you're scaling up (which will provide more memory), it's good practice to understand your current memory utilization patterns.
-
CloudWatch Metrics Baseline: Capture baseline metrics before the upgrade for post-upgrade comparison:
- CPU Utilization
- FreeableMemory
- DatabaseConnections
- Read/Write IOPS and Latency
- Read/Write Throughput
- BufferCacheHitRatio
- Commit Latency and Throughput
-
Backup Verification: Ensure automated backups are enabled and take a manual snapshot before proceeding with the modification.
-
Testing: If possible, test the instance class change in a non-production environment first to validate application performance with the new instance size.
Since you have downtime available, you can schedule the modification during your maintenance window. The upgrade process will involve a brief outage while the instance class is changed.
Best regards
Sources
How to perform an in-place upgrade - Amazon Aurora
Mastering PostgreSQL Upgrades: Your Ultimate RDS/Aurora Handbook | AWS re:Post
Relevant content
asked 3 years ago
