cert-manager — Certificate stuck in "Issued=False"

Symptom

kubectl get certificate shows READY=False with conditions like:

  • TLS handshake timeout
  • DNS challenge failed
  • Connection refused or i/o timeout

Root Cause

cert-manager cannot complete the ACME challenge (HTTP-01 or DNS-01) because:

Cause Scenario
DNS propagation delay The DNS records for the domain have not propagated before the challenge check
Firewall blocking port 80/443 HTTP-01 challenge requires inbound access to port 80 or 443 from the internet
A/AAAA records point to wrong IP The domain resolves to an IP that does not route to the ingress controller
Rate limiting Let's Encrypt production issuer has rate limits per domain per week

Fix

1. Use the Let's Encrypt staging issuer for testing

Create a staging issuer to avoid rate limits during testing:

# letsencrypt-staging.yaml
apiVersion: cert-manager.io/v1
kind: ClusterIssuer
metadata:
  name: letsencrypt-staging
spec:
  acme:
    server: https://acme-staging-v02.api.letsencrypt.org/directory
    email: [email protected]
    privateKeySecretRef:
      name: letsencrypt-staging-private-key
    solvers:
      - http01:
          ingress:
            class: nginx

Apply it:

kubectl apply -f letsencrypt-staging.yaml

Reference the staging issuer in your Certificate or ingress annotation:

cert-manager.io/cluster-issuer: letsencrypt-staging

2. Verify DNS records

dig +short A <your-domain>
dig +short AAAA <your-domain>    # only if you serve IPv6

Ensure the resolved IP matches your cluster's ingress controller external IP.

3. Check port reachability

# From an external machine
curl -v http://<your-domain>/.well-known/acme-challenge/test
curl -v https://<your-domain>

Port 80 (HTTP-01 challenge) and port 443 (TLS) must be reachable from the public internet. Check cloud firewall / security group rules and your local firewall.

4. Check the certificate request logs

kubectl describe certificate <cert-name> -n <namespace>
kubectl describe order -n <namespace>
kubectl describe challenge -n <namespace>
kubectl logs -l app=cert-manager -n cert-manager --tail=50

The logs will show exactly which step failed and why.

5. Switch to production issuer after testing

Once staging works, update to the production issuer:

server: https://acme-v02.api.letsencrypt.org/directory

Verification

kubectl get certificate <cert-name> -n <namespace> -w

Expected output — READY changes to True.

openssl s_client -connect <your-domain>:443 -servername <your-domain> 2>/dev/null \
  | openssl x509 -noout -issuer

Expected output — shows Let's Encrypt as the issuer.