The Tool Desk
Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →Changing an Amazon EKS add-on version does not upgrade your worker nodes. AWS documents add-on updates and node-group updates as separate operations. You can aim for uninterrupted service while replacing nodes, but no upgrade path guarantees zero downtime: the outcome depends on available capacity, pod scheduling, disruption budgets, and the path you choose.
Is there a “bridge add-on version trick” for upgrading EKS nodes?
AWS’s documented upgrade guidance does not describe changing an add-on version as a bridge for upgrading nodes. The control plane, data-plane nodes, and add-ons have distinct update operations. Add-ons do not automatically change when you upgrade a cluster’s Kubernetes minor version; you initiate an add-on update after checking compatibility. An add-on update therefore cannot stand in for a node upgrade.
As an Amazon Associate I earn from qualifying purchases.
AWS recommends a coordinated sequence: prepare and check readiness, upgrade the control plane, upgrade nodes, then review and update add-ons and custom controllers as needed. Keep client tools such as kubectl compatible with the cluster, and follow the version-skew guidance for your node type and Kubernetes version. See AWS best practices for EKS cluster upgrades and the EKS add-on update guide.
Do these 3 things before closing this tab:
1Repair Windows errors before they cause bigger problems2Fix the driver behind crashes, sound loss and screen glitches3Clear out junk files and repair common Windows errorsChoose a node replacement path
The right process depends on whether AWS manages the node group, you manage the nodes yourself, or you use EKS Auto Mode. These approaches differ in who replaces nodes and how workloads are moved.
#1 Best Overall
| Path | How nodes and pods are handled | Disruption-budget behavior | Recovery considerations |
|---|---|---|---|
| Managed node group | AWS performs the update and drains pods before terminating nodes. | A PodDisruptionBudget (PDB) can block draining and cause the update to fail. | The update API offers a force option for a PDB-related drain failure, but overriding a PDB can remove intended protection. Rolling back a node group to an earlier Kubernetes or AMI version is not supported. |
| Self-managed nodes | Create a replacement node group, allow it to take workloads, then taint and drain the old group to migrate pods. | Plan migration and workload capacity around your disruption budgets and scheduling constraints. | AWS describes migration to a new node group as a more graceful option than changing the AMI on the existing stack. |
| EKS Auto Mode | After a control-plane version update, Auto Mode incrementally replaces nodes. | Auto Mode observes PDBs during replacement. | Confirm which components Auto Mode manages and which add-ons or controllers remain your responsibility. |
For details, use AWS’s managed node group update API, self-managed node update guidance, and EKS Auto Mode version update guidance.
How to reduce disruption during an upgrade
Before updating a managed node group
- Check that workloads have enough replicas and that the cluster has capacity to schedule replacement pods.
- Review PDBs and scheduling constraints that could prevent pods from leaving a node or being placed elsewhere.
- Follow the supported control-plane and node version path for your Kubernetes version and node type.
- Do not use the force option as a routine way to make an update complete. If a PDB blocks draining, forcing the update may defeat the protection that budget was intended to provide.
When replacing self-managed nodes
- Create a new node group with the intended node configuration.
- Let the new group join the cluster and confirm it can accept the workloads.
- Taint and drain the old group, migrating workloads to the replacement nodes.
- Retire the old group only after workloads have moved successfully.
AWS identifies this replacement-group approach as more graceful than changing the AMI on the existing self-managed stack. The exact scheduling and migration outcome still depends on your workloads and cluster capacity.
When using EKS Auto Mode
Auto Mode incrementally replaces nodes after the control-plane update and observes PDBs. Check which infrastructure and components it manages in your cluster; update any self-managed add-ons or controllers that remain outside that managed scope.
Check add-on compatibility separately
After a Kubernetes version change, review each add-on and its compatibility with the cluster version, then initiate an update when appropriate. Do not assume a cluster upgrade also updates add-ons.
Rank #3
CoreDNS deserves a specific check: add-on versions can include PDB configuration, and an existing PDB or custom Corefile or readiness settings may affect the upgrade. Review the applicable AWS CoreDNS management guidance before changing its add-on version.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.What “without downtime” can realistically mean
Node replacement can preserve service when workloads remain available on other schedulable nodes, but “zero downtime” is not assured by an add-on change or by any one node-upgrade option. A PDB may prevent a managed-node drain; forcing the update can bypass that protection. For self-managed nodes, migrating onto a ready replacement group gives you a more controlled transition, but it still relies on capacity and workload behavior. Treat availability as a goal to validate for your own cluster, not a guarantee built into the upgrade mechanism.
Quick Recap
Best Value
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
