# Self-Healing CI: Recovering a Slow CDN Purge

> A post-deploy check that fails because a CDN purge had not propagated hit a timing issue, not a deploy bug. See the manual fix and how self-healing CI waits and retries.

Source: https://latchkey.dev/learn/self-healing-ci/self-healing-cdn-purge-slow  
Updated: 2026-06-26

A post-deploy check that still sees stale content was usually just ahead of CDN propagation -- waiting for the purge to finish makes the check pass.

## What makes a failure safely retryable

Automatic retry is only correct for failures that are genuinely transient. Retrying a deterministic failure wastes minutes and hides a real defect, so the classification matters more than the retry mechanism.

- Safe to retry: network timeouts, registry 5xx, transient DNS failures, a service container that was not ready, a spot instance reclaimed mid-run.
- Not safe to retry: assertion failures, compile errors, lint violations, anything that fails identically on every attempt.
- Ambiguous, and worth investigating rather than retrying: out-of-memory kills, disk exhaustion, and flaky tests. These repeat under load and a retry only hides the trend.
- Always record that a retry happened. A pipeline that silently retries is a pipeline whose real failure rate you do not know.

## FAQ

### What causes Self-Healing CI: recovering a slow CDN purge?

A post-deploy verification fails because a CDN cache purge or invalidation had not finished propagating, so the check still saw stale content. The deploy and the purge request both succeeded; propagation across edge nodes simply lagged. A human re-runs the check after a short wait and it passes unchanged.

### How do I fix Self-Healing CI: recovering a slow CDN purge manually?

[object Object]

### Can Self-Healing CI: recovering a slow CDN purge be fixed automatically?

A propagation lag is distinguishable from a failed deploy by whether the fresh content appears shortly after. A self-healing CI pipeline detects the stale-content condition, waits for the purge to propagate and re-checks, and only escalates if the content never updates, which is the real signal of a deploy problem rather than slow

---

Latchkey runs CI/CD that repairs its own failures. Agent entry points: https://latchkey.dev/agent.txt, https://latchkey.dev/openapi.json, https://latchkey.dev/llms.txt