Zero-Downtime DNS Cutover & SSL Checklist for High-Traffic E-Commerce Launches
The senior DevOps runbook: executing zero-downtime DNS migrations, pre-warming SSL certificates, and preventing lost cart transactions during cutovers.
In digital agency operations and white-label engineering, launch day is the moment of truth. Months of meticulous design, backend programming, and testing culminate in a single high-stakes event: pointing live production DNS to the new platform.
For high-traffic e-commerce brands generating thousands of euros per hour, an amateur cutover is catastrophic:
- Global DNS resolvers take 24 to 48 hours to propagate, leaving half the world seeing the old website.
- New visitors encounter terrifying browser SSL security warnings (
NET::ERR_CERT_COMMON_NAME_INVALID) because certificates were not pre-provisioned. - Customers place orders on the legacy store while others order on the new store, creating a nightmarish split-brain inventory and accounting mess.
- The client panics, the agency leadership scrambles, and trust is severely damaged.
Zero-downtime DNS cutovers are not achieved by luck or hoping for fast propagation; they are the result of executing a disciplined, phase-gated DevOps runbook.
Below is our white-label agency checklist: detailing the exact chronological sequence of DNS, SSL, database synchronization, and edge cache steps required to execute flawless cutovers with zero minutes of customer-facing downtime.
1. Phase 1: 72 Hours Prior to Cutover (The TTL Drop)
The most common cause of prolonged DNS propagation delays is high Time to Live (TTL) values on existing DNS records.
Standard DNS records are frequently configured with a TTL of 86,400 seconds (24 hours) or 14,400 seconds (4 hours). If you change an A record with a 24-hour TTL, ISPs across the world will continue routing traffic to the old IP address for a full day.
[72 Hours Before Launch: Lowering TTL]
├── Existing A Record: `example.com` ──► TTL: 86400s (24 Hours)
└── ACTION: Lower TTL to 300s (5 Minutes)
│
▼
(Wait 72 hours so all global resolvers flush old 24h caches)
│
▼
[Launch Window: Instant Record Switching within 300 Seconds]
72-Hour Checklist Items:
- Lower TTL on Primary Records: Reduce TTL on
@(root A record) andwww(CNAME) to300seconds (5 minutes). - Audit MX and SPF Records: Ensure email routing records (Google Workspace, Microsoft 365) are documented and untouched during the web cutover.
- Verify Staging Host Header Routing: Ensure the new server is configured to accept requests with the production host header (
Host: example.com) rather than only staging subdomains.
2. Phase 2: 24 Hours Prior to Cutover (Pre-Warming SSL)
Never wait until DNS has been flipped to generate your production SSL/TLS certificate with Let’s Encrypt or your cloud CDN. If the automated HTTP-01 challenge fails during launch, your new store will immediately display SSL warning banners.
┌────────────────────────────────────────────────────────────────────────┐
│ Pre-Provisioning Production SSL/TLS │
├───────────────────────────────────┬────────────────────────────────────┤
│ ❌ Reactive HTTP-01 Validation │ ✅ Proactive DNS-01 Pre-Warming │
├───────────────────────────────────┼────────────────────────────────────┤
│ 1. Switch DNS to new IP │ 1. Add TXT verification record to │
│ 2. Let's Encrypt queries domain │ existing authoritative DNS │
│ 3. If propagation lags, SSL fails │ 2. CA validates domain control │
│ 4. Site shows "Insecure" warning │ 3. SSL certificate issued in RAM │
│ 5. Shoppers abandon cart │ 4. Zero seconds of SSL downtime │
└───────────────────────────────────┴────────────────────────────────────┘
Using Cloudflare Custom Hostnames or Let’s Encrypt DNS-01 challenges, generate and deploy your production SSL certificate 24 hours in advance while the old store is still handling 100% of live traffic.
3. Phase 3: The 30-Minute Cutover Window (Execution Steps)
Schedule the cutover during the merchant’s historical lowest traffic window (typically Tuesday or Wednesday between 02:00 and 04:00 AM local time).
[02:00 AM] Freeze Legacy Orders: Place old store in Maintenance Mode
│
[02:05 AM] Final Database Delta Sync: Export incremental orders & customers
│
[02:15 AM] Import Delta to New Production Database & Verify Row Counts
│
[02:20 AM] Flip DNS A/CNAME Records to New Infrastructure (TTL: 300s)
│
[02:25 AM] Run End-to-End Live Checkout Smoke Test (Real Credit Card)
│
[02:30 AM] Disable Maintenance Mode ──► CUTOVER COMPLETE!
Step 1: Prevent Split-Brain Transactions
Place the legacy e-commerce store into maintenance mode or disable the checkout button with an informative banner. This prevents new orders from being written to the old database during the final synchronization.
Step 2: Database Delta Migration
Export all orders, customers, and inventory changes created on the old store since the staging database snapshot was taken. Import the delta into the new database and confirm that latest order IDs match perfectly.
Step 3: Switch DNS Records
Update the A record to the new server IP or point the CNAME to Cloudflare/CloudFront. Because TTL was reduced to 300 seconds 72 hours earlier, global traffic switches over to the new store within 5 minutes.
Step 4: Live Checkout Smoke Test
Using a real credit card, execute a live purchase on the production domain. Verify payment gateway authorization, order confirmation email receipt, and webhook dispatch to fulfillment systems.
4. Phase 4: Post-Launch Hardening (Restoring TTL)
Once the new platform is operating stably and post-launch analytics confirm normal traffic flow:
- Restore Standard TTL: Increase TTL back to
14400(4 hours) or86400(24 hours) to reduce DNS resolver lookup latency and protect authoritative nameservers from excessive query volume. - Decommission Staging Access: Revoke public access to staging subdomains and update robots.txt to prevent search engine indexing of internal build URLs.
- Archive Legacy Database: Create an encrypted full backup of the legacy store database and store it in cold cloud storage before shutting down the old server.
5. Summary: Flawless Launches Build Agency Reputations
An e-commerce launch is remembered not by the months of smooth design sprints that preceded it, but by whether the cutover succeeded without downtime.
By following this phase-gated runbook:
- You eliminate propagation delays by pre-emptively lowering TTL.
- You avoid browser security warnings by pre-provisioning SSL certificates.
- You protect transaction integrity by preventing split-brain database states.
Your agency executes frictionless, zero-downtime launches that impress enterprise clients and protect retail revenues.
To explore related e-commerce deployment protocols and white-label management standards:
- Read our foundational playbook The Agency Leader’s Playbook to White-Label Web Engineering.
- Study high-load architecture in Hardened WordPress: Bedrock, Sage & WooCommerce Guide.
- Review database maintenance in Taming wp_actionscheduler & wp_postmeta Bloat.
Ready to ensure your next major e-commerce migration launches with zero downtime? Contact our white-label engineering team to manage your technical cutovers.
Frequently Asked Questions
How can an engineering team prevent DNS propagation downtime during a domain cutover?→
How do you avoid SSL handshake errors when switching to a new server or CDN?→
How do you prevent split-brain orders between old and new stores during cutover?→
Scaling an agency? Offload your engineering bottlenecks.
Fonda Digital provides senior white-label full-stack development, headless migrations, and hardened WordPress builds under bilateral NDA. Predictable wholesale margins.