RDS MySQL 8.0 Has Been Billing Extended Support Since August: Find It and Price It

If an RDS for MySQL 8.0 instance is still running, its bill changed on August 1, 2026. RDS ended standard support for 8.0 on July 31 and now bills Extended Support by vCPU-hour in addition to the DB instance, storage, backup, and transfer charges. The easy mistake is treating that as an upgrade decision for later. It is already a recurring line item.
This post gives you a per-Region inventory, a billing-data path, a cost model, and the decision points for an 8.0 to 8.4 upgrade. The same inventory also exposes older PostgreSQL and Aurora versions that have crossed their own support boundaries.
TL;DR
- RDS MySQL 8.0 entered paid Extended Support on August 1, 2026. Year 3 pricing begins August 1, 2028, and support ends July 31, 2029. AWS's MySQL release calendar also says MySQL 5.7 and 8.0 are now only available under Extended Support.
- Use
EngineLifecycleSupportto see enrollment, but combine it with engine version and support dates. Enrollment alone does not mean a charge until standard support has ended. - RDS charges provisioned MySQL Extended Support per vCPU-hour. AWS says it also applies to Multi-AZ standby instances and to read replicas on the same version, so count every one.
- A new CLI or API-created instance defaults to Extended Support enabled if you omit the setting. Specify
open-source-rds-extended-support-disabledif automatic upgrade or creation failure is preferable to silent enrollment. - EKS has the same financial shape at the control-plane level: $0.10 per cluster-hour during standard support and $0.60 in extended support. Set its upgrade policy to
STANDARDonly when you accept AWS automatically upgrading it at the end of standard support.
Prerequisites
- AWS CLI v2 credentials able to call RDS, EC2, EKS, and Cost Explorer in the intended accounts.
- Billing permission for
ce:GetCostAndUsage. For organization-wide results, run it from the AWS Organizations management account or a role there; member accounts see only their own Cost Explorer data. - A current RDS pricing view for each Region you operate in. Rates and support calendars are Region and date sensitive.
- An upgrade test environment and an application owner before changing an engine version or lifecycle setting.
What changed, and what RDS is charging
AWS lists these RDS for MySQL dates: MySQL 8.0 community end of life was April 30, 2026, RDS standard support ended July 31, 2026, Year 1 pricing began August 1, 2026, Year 3 begins August 1, 2028, and Extended Support ends July 31, 2029. MySQL 5.7's calendar is already in Year 3: it began charging March 1, 2024, stepped up March 1, 2026, and ends June 30, 2029. MySQL 8.4 remains in standard support until July 31, 2029.
The other calendars deserve the same review. RDS for PostgreSQL 11, 12, and 13 started paid Extended Support on April 1, 2024, March 1, 2025, and March 1, 2026. PostgreSQL 14 reaches the boundary after standard support ends February 28, 2027, with Year 1 pricing from March 1, 2027. Those dates come from the RDS PostgreSQL release calendar. Aurora has a different schedule: Aurora MySQL 3, compatible with MySQL 8.0, remains in standard support until April 30, 2028, while Aurora PostgreSQL's calendar shows the same paid dates as RDS PostgreSQL for 11 through 14. Do not apply the RDS MySQL 8.0 deadline to Aurora MySQL 3.
For provisioned RDS for MySQL, the pricing page describes Extended Support as per vCPU per hour. The vCPU count is the instance class's published vCPU count, not the CPU percentage you happened to use. AWS's DB instance class tables are the source of record for the class. The charge applies to Multi-AZ standby instances, according to the Extended Support charges guide, so a traditional Multi-AZ deployment has more billable vCPUs than its primary alone. Include every read replica too: the pricing page says replicas running a version past standard support are charged as well, each with its own class and vCPU count.
AWS documents Aurora Serverless v2 capacity in ACUs, but the RDS MySQL pricing page's vCPU formula is for provisioned RDS for MySQL. Do not convert ACUs to vCPUs for an invoice forecast unless the Aurora price page for the exact engine and Region supplies that rate.
Inventory every Region before estimating
Start with a read-only snapshot of RDS instances and clusters. EngineLifecycleSupport reports enrollment using open-source-rds-extended-support or open-source-rds-extended-support-disabled; it is not a substitute for checking the version.
for region in $(aws ec2 describe-regions \
--query 'Regions[].RegionName' --output text --no-cli-pager); do
echo "### $region: DB instances"
aws rds describe-db-instances --region "$region" \
--query 'DBInstances[].{Id:DBInstanceIdentifier,Engine:Engine,Version:EngineVersion,Class:DBInstanceClass,MultiAZ:MultiAZ,Source:ReadReplicaSourceDBInstanceIdentifier,Support:EngineLifecycleSupport}' \
--output table --no-cli-pager
echo "### $region: clusters, including Aurora"
aws rds describe-db-clusters --region "$region" \
--query 'DBClusters[].{Id:DBClusterIdentifier,Engine:Engine,Version:EngineVersion,Mode:EngineMode,Members:DBClusterMembers[].DBInstanceIdentifier,Support:EngineLifecycleSupport}' \
--output table --no-cli-pager
done
describe-db-clusters covers Aurora and Multi-AZ DB clusters. For Aurora, lifecycle setting lives at the cluster level, then the Members list tells you which DB instances to count. The RDS creation documentation confirms that members of a cluster share this setting.
Use the service's support-date response rather than copying a calendar into automation:
aws rds describe-db-major-engine-versions \
--engine mysql \
--major-engine-version 8.0 \
--query 'DBMajorEngineVersions[].{Engine:Engine,Major:MajorEngineVersion,Lifecycles:SupportedEngineLifecycles}' \
--output json --no-cli-pager
The support-date API guide documents SupportedEngineLifecycles, including its support name and start and end timestamps. Repeat it for postgres, aurora-mysql, and aurora-postgresql where those engines are in scope.
Find the charge in Cost Explorer
First discover the usage-type values in your bill, and do not hard-code a filter before you have seen the returned values. This query groups all RDS usage by usage type, then jq narrows the returned group names locally:
aws ce get-cost-and-usage \
--time-period Start=2026-10-01,End=2026-11-01 \
--granularity MONTHLY \
--metrics UnblendedCost UsageQuantity \
--filter '{"Dimensions":{"Key":"SERVICE","Values":["Amazon Relational Database Service"]}}' \
--group-by Type=DIMENSION,Key=USAGE_TYPE \
--output json \
| jq '.ResultsByTime[].Groups[]
| select(.Keys[0] | test("extended|support"; "i"))'
Inspect the result before adding a Cost Category, budget, CUR query, or saved Cost Explorer filter. The point is to prove the actual usage type in the billing account, not to make a dashboard look complete with a guessed string. For detailed allocation, query the cost and usage data export using the service, usage-type, usage amount, and cost columns that appear in your exported schema. Data Exports schemas vary by export choice, so copy column names from the table's own definition rather than borrowing an Athena query from a different export. Pair this with cost-allocation tags if a team needs to own the upgrade decision.
Put a monthly number on MySQL 8.0
The formula is simple once the inventory is right:
Extended Support per month = billable vCPUs x billed hours x rate per vCPU-hour
As checked November 11, 2026, AWS's RDS for MySQL pricing page shows $0.100 per vCPU-hour for Years 1 and 2 and $0.200 for Year 3 in its US East rate table, and its FAQ example for a US East (N. Virginia) db.r5.large uses the same $0.100 Year 1 rate. AWS says the actual price depends on MySQL version, Region, and calendar year, so confirm the selected Region and engine in the current pricing view before treating this as a purchase forecast.
For a transparent planning example, assume two single-AZ db.r6g.2xlarge MySQL 8.0 instances in US East (N. Virginia), 8 vCPUs each, at 730 hours per month. That is 16 billable vCPUs. The Extended Support addition is 16 x 730 x $0.100 = $1,168 in Years 1 and 2. At the Year 3 rate it is $2,336. This excludes the instance's normal price, storage, I/O, backups, and transfer. Add the standby vCPUs before using this calculation for Multi-AZ.
EKS has a smaller, separate support timer
EKS charges per cluster-hour, not per vCPU. AWS lists $0.10 per cluster-hour during standard support and $0.60 in extended support. A Kubernetes version gets 14 months of standard support and 12 months of extended support. Check each cluster's version and policy:
aws eks describe-cluster --name platform-prod --region us-east-1 \
--query 'cluster.{version:version,support:upgradePolicy.supportType}' \
--output table --no-cli-pager
The supportType field is STANDARD or EXTENDED. AWS's upgrade policy guide says STANDARD causes an automatic upgrade at the end of standard support, while EXTENDED keeps the version in extended support. Change it only after validating the next Kubernetes version:
aws eks update-cluster-config --name platform-prod --region us-east-1 \
--upgrade-policy supportType=STANDARD \
--no-cli-pager
The EKS pricing page has the current rates. This setting avoids the extended-support choice by accepting automatic control-plane version movement, not by making an old version free.
Decide whether to upgrade or pay knowingly
For RDS MySQL, 8.0 to 8.4 is a supported major upgrade path. Outside the lifecycle behavior described below, AWS does not automatically apply major upgrades, because they can break compatibility. Its major-upgrade guide says RDS runs mandatory 8.0 to 8.4 prechecks before stopping the instance. Failures are logged in PrePatchCompatibility.log, so treat that log as an upgrade backlog, not a reason to ignore the charge.
Test three classes of problems before a production cutover:
mysql_native_passwordis disabled by default in MySQL 8.4, so accounts and old clients that rely on it need a deliberate plan. MySQL's documentation describes the change and its connection failures.- MySQL 8.4 changes defaults including
innodb_adaptive_hash_indexfromONtoOFF,innodb_change_bufferingfromalltonone, andinnodb_io_capacityfrom 200 to 10000. Review parameter groups and performance expectations against MySQL's changed-defaults table. - RDS prechecks reject obsolete types or functions, invalid trigger definers or creation context, new reserved-word collisions, overlapping
mysqlsystem-table names, and obsoletesql_modesettings.
For a production path, use an RDS Blue/Green deployment when its prerequisites fit the database. It creates a green staging copy that stays in sync with blue, permits testing before cutover, and AWS says switchover is usually under one minute, although workload conditions can make it longer. See the RDS Blue/Green overview and creation guide. The method is supported for RDS MySQL 8.0 and 8.4. It is a better place to find application and parameter incompatibilities than the maintenance window.
Disabling Extended Support on an existing instance or cluster already past standard support does not leave it on the old version without a charge. AWS states that it automatically upgrades to the next supported major version. Treat that change as an upgrade operation, with testing and an owner, not as a FinOps toggle.
Prevent the next surprise
For new CLI-created RDS DB instances and Multi-AZ DB clusters, omitting lifecycle support defaults to enabled. Add the explicit opt-out to creation automation:
aws rds create-db-instance \
--db-instance-identifier app-mysql \
--engine mysql \
--engine-version 8.4 \
--db-instance-class db.r6g.large \
--allocated-storage 100 \
--master-username appadmin \
--manage-master-user-password \
--no-publicly-accessible \
--db-subnet-group-name app-private \
--vpc-security-group-ids sg-0123456789abcdef0 \
--engine-lifecycle-support open-source-rds-extended-support-disabled \
--no-cli-pager
After a support boundary has passed, that setting makes a new instance creation fail rather than enroll it. Before the boundary, a disabled instance is automatically upgraded on or shortly after the end of standard support. The RDS lifecycle documentation defines both behaviors.
Terraform exposes the equivalent engine_lifecycle_support argument on aws_db_instance and aws_rds_cluster. Make it explicit in modules, then add a policy check that rejects the enabled value where the organization expects an upgrade instead. The site's Terraform best-practices guide has useful patterns for keeping that module input visible in review. For enforcement outside Terraform, use a custom AWS Config rule that you have tested in your own account.
The decision record should have three fields: the affected version, the monthly Extended Support addition, and either a named upgrade date or an approved end date for paid support. That turns an automatic charge into a time-bounded engineering choice. If you need a rehearsal pattern for the release itself, the site's blue-green deployment guide is a useful companion.
Try it hands-on
Run the commands from this article in the browser. Nothing to install.
AWS VPC Networking Simulator
Learn AWS networking fundamentals with an interactive VPC simulator. Visualize how traffic flows through public and private subnets, understand NAT Gateways, Internet Gateways, and route tables.
PostgreSQL Terminal Simulator
Practice the psql tool and Postgres engine in an interactive browser terminal. Explore with meta-commands, read real EXPLAIN plans, add an index to turn a sequential scan into a bitmap index scan, and use BEGIN and ROLLBACK to make changes safe, all against a small e-commerce database with verified PostgreSQL output.
We earn commissions when you shop through the links below.
Bitrise
Managed runners for GitHub Actions
Bitrise Build Hub runs your GitHub Actions jobs on M4 Macs and AMD Linux machines. Change runs-on and keep your workflows, secrets and logs on GitHub.
Atomsized
AWS platform engineering and GitOps
Design and automation for reliable AWS and Kubernetes platforms, safer delivery workflows, and preview and UAT environments your engineers can understand and own.
DigitalOcean
Cloud infrastructure for developers
Simple, reliable cloud computing designed for developers
DevDojo
Developer community & tools
Join a community of developers sharing knowledge and tools
SMTPfast
Developer-first email API
Send transactional and marketing email through a clean REST API. Detailed logs, webhooks, and embeddable signup forms in one dashboard.
QuizAPI
Developer-first quiz platform
Build, generate, and embed quizzes with a powerful REST API. AI-powered question generation and live multiplayer.
Want to support DevOps Daily and reach thousands of developers?
Become a SponsorTags
Found an issue?
Related Posts
Also worth your time on this topic
Your Cloud Bill Is an Organizational Problem, Not a Technical One
Why your AWS bill keeps growing despite cost optimization efforts. The real driver is team structure, not instance size. A framework for tying cloud spend to service ownership and engineering accountability.
Cost Allocation Tags Basics
Can you explain what cost allocation tags are and why teams use them across AWS, GCP, and Azure?
junior
AWS Security Checklist
Essential security configuration checklist for AWS cloud environments.
45-60 minutes