Skip to content

fix: switch HA NLB to internal scheme to fix hairpin routing #760

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
ArangoGutierrez merged 1 commit into from
Mar 31, 2026
Merged

fix: switch HA NLB to internal scheme to fix hairpin routing #760

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
7 changes: 4 additions & 3 deletions pkg/provider/aws/nlb.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -54,15 +54,16 @@ func (p *Provider) createNLB(cache *ClusterCache) error {
}
lbName := nlbBaseName + nlbSuffix

// Use the public subnet for the internet-facing NLB
// Use the public subnet for the internal NLB (same subnet as instances)
subnetIDs := []string{cache.PublicSubnetid}

// Create load balancer
// Create load balancer — internal scheme avoids hairpin routing issues
// where nodes connecting to the NLB's public IP get i/o timeouts
createLBInput := &elasticloadbalancingv2.CreateLoadBalancerInput{
Name: aws.String(lbName),
Type: lbType,
Subnets: subnetIDs,
Scheme: elbv2types.LoadBalancerSchemeEnumInternetFacing,
Scheme: elbv2types.LoadBalancerSchemeEnumInternal,
IpAddressType: elbv2types.IpAddressTypeIpv4,
Tags: p.convertTagsToELBv2Tags(),
}
Expand Down
17 changes: 0 additions & 17 deletions pkg/provisioner/templates/kubeadm_cluster.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -153,23 +153,6 @@ fi

# Initialize cluster
if [[ ! -f /etc/kubernetes/admin.conf ]]; then
# Wait for control-plane endpoint to be resolvable (NLB DNS may take time)
if [[ "$CONTROL_PLANE_ENDPOINT" == *"elb.amazonaws.com"* ]] || \
[[ "$CONTROL_PLANE_ENDPOINT" == *"amazonaws.com"* ]]; then
holodeck_log "INFO" "$COMPONENT" "Waiting for NLB DNS to resolve: ${CONTROL_PLANE_ENDPOINT}"
for i in {1..30}; do
if host "${CONTROL_PLANE_ENDPOINT}" &>/dev/null || \
getent hosts "${CONTROL_PLANE_ENDPOINT}" &>/dev/null; then
holodeck_log "INFO" "$COMPONENT" "NLB DNS resolved successfully"
break
fi
if [[ $i -eq 30 ]]; then
holodeck_log "WARN" "$COMPONENT" "NLB DNS not yet resolved after 5 min, proceeding anyway"
fi
sleep 10
done
fi

INIT_ARGS=(
--kubernetes-version="${K8S_VERSION}"
--pod-network-cidr=192.168.0.0/16
Expand Down
Loading