docs: Clarify expireAfter behavior and interaction with disruption budgets

[moko-poi]

How can the docs be improved?

The current Expiration documentation does not clearly explain the relationship between expireAfter and other disruption reasons (Drift, Consolidation, Emptiness).

Current Problem

Users may misunderstand that expireAfter: 720h (30 days) guarantees nodes will live for at least 30 days. However, in reality:

• expireAfter is a maximum node lifetime (ceiling), not a minimum guarantee
• Other disruption reasons (e.g., Drifted, Consolidation) can terminate nodes earlier if disruption budgets allow
• This can lead to unexpected node terminations when combined with disruption budget schedules

Proposed Changes

The Expiration section in the karpenter.sh documentation should clarify:

1. expireAfter defines the maximum node lifetime, not a guaranteed minimum
2. Nodes can be disrupted earlier by other reasons (Drift, Consolidation, Emptiness) if budgets permit
3. To enforce a minimum node lifetime, users need to configure appropriate disruption budgets

Target Files

• website/content/*/concepts/disruption.md or similar files in the Expiration/Disruption documentation

Why are we solving this issue?

To prevent user confusion about the expireAfter behavior and help them properly configure node lifecycle management.

Reference

• Related to upstream issue: Document expireAfter in more detail kubernetes-sigs/karpenter#2530
• Background: Expiration changed from graceful to forceful in v1 (PR Skip provisioner.spec.securityGroupSelector if provided in custom launch template #1303)

---

• Please vote on this issue by adding a 👍 reaction to the original issue to help the community and maintainers prioritize this request
• Please do not leave "+1" or "me too" comments, they generate extra noise for issue followers and do not help prioritize the request
• If you are interested in working on this issue or have submitted a pull request, please leave a comment