External Hyperstream (HS) Launch — DNS, Certificates, Load Balancer, and FQDN Updates

External Hyperstream (HS) Launch — DNS, Certificates, Load Balancer, and FQDN Updates

External Hyperstream (HS) Launch — DNS, Certificates, Load Balancer, and FQDN Updates

This page is a child of the Deployment Guide and provides end‑to‑end steps to enable External HS‑Hyperstream launching. It complements the pages: Prerequisites, https://apportoteam.atlassian.net/wiki/spaces/PUB/pages/1368817686, and Running the on-prem cluster install wizard.

Overview

To allow external client endpoints to launch Hyperstream sessions through your network perimeter, you must configure public DNS, TLS certificates, a perimeter load balancer or gateway (and optional HA), and verify/update Hyperstream FQDNs post‑install.

  • Step 1 — Create External DNS A record

  • Step 2 — Obtain and prepare a third‑party wildcard certificate

  • Step 3 — Update the load balancer (NAT, SSL termination, backends, VIP/HA)

  • Step 4 — Update Hyperstream FQDN in Apporto (Resource Hub and HS node)

Step 1 — Create an External DNS A Record

Create a public DNS A record pointing to your external, internet‑reachable IP (often a NAT IP). Use your preferred DNS provider (e.g., Cloudflare, Route53, GoDaddy).

  • Choose the host name to match your external HS URL, for example: hs.example.com.

  • Map the A record to your external public IP (e.g., 50.50.50.50).

  • If using a provider with proxy/CDN features, ensure end‑to‑end TLS and WebSockets are supported for HS traffic (disable proxy if it interferes with WebSockets).

Host: hs.example.com Type: A Value: 50.50.50.50 TTL: Automatic or 300s

Step 2 — Obtain and Prepare a Third‑Party Wildcard Certificate

Acquire a trusted wildcard certificate that covers your HS hostname (e.g., *.example.com) from a public CA such as Sectigo or Let’s Encrypt. You will install this on your perimeter load balancer or gateway.

  • Certificate must include: server certificate, any intermediate CA certificates, and the private key.

  • Key size: 2048‑bit or stronger. Use modern TLS ciphers on the LB/gateway.

  • SAN should include the exact HS FQDN (most wildcard certs already match).

If using HAProxy, create a single PEM file containing the full chain and key in this order:

  1. server.crt (leaf)

  2. intermediate.crt (one or more)

  3. server.key (private key)

cat server.crt intermediate.crt server.key > /etc/haproxy/certs/hs-example-com.pem chmod 600 /etc/haproxy/certs/hs-example-com.pem chown root:root /etc/haproxy/certs/hs-example-com.pem

https://apportoteam.atlassian.net/wiki/spaces/PUB/pages/1368817686

Step 3 — Update the Load Balancer

Terminate HTTPS on your perimeter device and forward HS traffic to the internal HS/gateway endpoints. Ensure WebSockets are supported end‑to‑end.

3.1 Public NAT and Port Mapping

  • Create a DNAT rule mapping external 443 to your internal LB/gateway VIP or node.

  • Example mapping: 50.50.50.50:443 → 10.200.0.70:443.

  • Allow inbound TCP/443 from the internet and health probes from your monitoring sources.

3.2 HAProxy reference configuration

The following is a minimal pattern. Adjust IPs, health checks, ciphers, and timeouts per your standards.

global log /dev/log local0 maxconn 5000 tune.ssl.default-dh-param 2048 defaults log global mode tcp option tcplog timeout connect 10s timeout client 90s timeout server 90s # HTTPS listener (terminate TLS and forward to HS/gateway over TLS or TCP) frontend fe_https_hs bind *:443 ssl crt /etc/haproxy/certs/hs-example-com.pem alpn h2,http/1.1 mode http option forwardfor http-response set-header Strict-Transport-Security "max-age=15552000; includeSubDomains" default_backend be_hs_nodes backend be_hs_nodes mode http option httpchk GET /health http-check expect status 200 balance roundrobin server hs-node-1 10.200.0.101:30443 check ssl verify none server hs-node-2 10.200.0.102:30443 check ssl verify none # If you use the Apporto Secure Gateway, point to gateway nodes on 443 instead of 30443

If your internal HS/gateway expects TLS, use ssl verify none during initial setup only. Replace with CA verification in production.

3.3 High availability (keepalived VIP)

For active/standby HAProxy VMs, configure a VRRP virtual IP (VIP) so the internal target is stable. Health checks should trigger failover.

vrrp_instance VI_1 { state MASTER interface ens160 virtual_router_id 51 priority 150 advert_int 1 authentication { auth_type PASS auth_pass change_me } virtual_ipaddress { 10.200.0.70/24 dev ens160 } }
Internet → LB: TCP/443 LB → HS/gateway nodes: TCP/30443 (direct HS), or TCP/443 (via Apporto Secure Gateway) VRRP between HA nodes: protocol 112 (IP protocol), allow between HAProxy peers

Step 4 — Update Hyperstream FQDN in Apporto

After completing the On‑Prem OVA cluster installation, verify and, if needed, update the external FQDN to match the DNS record created in Step 1.

  1. Open Apporto Admin Portal → Resource Hub.

  2. Locate the Hyperstream component and the Secure Gateway (if used).

  3. Set the Hyperstream External URL/FQDN to your public hostname (e.g., https://hs.example.com).

  4. If using an Apporto Secure Gateway, set its public FQDN similarly and ensure the LB routes are correct.

  5. Save changes and validate connectivity.

Do not modify the “Secret” value shown for Hyperstream or related resources unless explicitly instructed by Apporto.

Validation

  • Launch a Hyperstream session from an external client and confirm the browser URL is your external FQDN.

  • Verify TLS padlock is valid (issued by your chosen CA, matches hostname, full chain).

  • Confirm session stability (WebSocket upgrades succeed; no mixed‑content or certificate warnings).

Troubleshooting

DNS A record resolves externally but not internally: consider split‑horizon DNS or ensure clients use public resolvers. Certificate chain errors: ensure intermediate certs are included and in correct order in the PEM. Hostname mismatch: the CN/SAN must match the exact external FQDN.
WebSockets failing: confirm LB supports HTTP/1.1 or ALPN with h2/http1.1 and preserves Upgrade/Connection headers if terminating HTTP. NAT in place but no response: verify security groups/firewalls on 443 and the internal 30443/443 path. HA failover unreliable: allow VRRP protocol 112 between HA peers; align priorities and health checks.

Change Log

  • Initial publication: May 21, 2026

Related Pages