CI "Temporary failure in name resolution" (EAI_AGAIN)

By Kaveh Alemi·Latchkey

EAI_AGAIN means the DNS resolver temporarily could not answer, a soft, retryable failure. In CI it usually reflects a momentary resolver or network hiccup, not a bad hostname.

What this error means

A step fails with Temporary failure in name resolution or getaddrinfo EAI_AGAIN <host>. Re-running the same job typically succeeds without any change.

shell

curl: (6) Could not resolve host: registry.example.com
Temporary failure in name resolution

Common causes

The DNS resolver was momentarily unavailable

A transient resolver timeout returns EAI_AGAIN; a retry a moment later resolves fine.

Network not fully up yet

Very early in a job or container start, networking/DNS may not be ready, causing a brief EAI_AGAIN window.

How to fix it

Retry the resolution-dependent step

A short backoff loop absorbs the transient resolver failure.

shell

for i in 1 2 3 4 5; do
  getent hosts registry.example.com && break
  sleep 3
done

Stabilize resolution

Ensure networking is up before the first lookup.
Verify resolver config (/etc/resolv.conf).
Use a reliable upstream resolver if you control the runner.

How to prevent it

Add retries around the first network calls in a job.
Avoid lookups before the network is ready.
On managed runners, EAI_AGAIN-class blips are detected and the job is automatically retried, so a one-off failure does not break the build.

Frequently asked questions

What causes "Temporary name resolution"?

A transient resolver timeout returns EAI_AGAIN; a retry a moment later resolves fine.

How do I fix Temporary name resolution?