Zero-Downtime Cloud Migration & Core Web Vitals Speed Overhaul
Softlimes migrated a portfolio of WordPress, WooCommerce, and Laravel websites from overcrowded shared cPanel servers to tuned Linux cloud instances running Nginx, PHP 8.x, Redis object caching, and automated off-site backups.
Many growing businesses keep their websites on the same shared hosting plan they bought five years ago. As traffic, WooCommerce orders, and database tables grow, Time to First Byte (TTFB) climbs past two seconds and plugin updates start timing out.
Moving to a faster server sounds risky when your business relies on live enquiries, active store checkouts, and domain email (Microsoft 365 or Google Workspace).
We executed a staged migration workflow: provisioning Nginx + PHP-FPM 8.x + Redis servers, testing full clones on staging hostnames, preserving all MX/SPF/DKIM email records, and running low-TTL DNS cutovers with final database delta syncs.
Challenge:
Escaping slow shared hosting without dropping emails or orders
The client sites suffered from three infrastructure bottlenecks on their old host:
- High server TTFB: Uncached WooCommerce and Laravel database queries took 1.8 to 3.2 seconds just to return the first byte of HTML.
- Same-disk backup risk: Backups were stored inside the same cPanel disk quota, slowing down the server and offering no protection if the host went down.
- Fear of migration downtime: Previous server moves had broken SPF/DKIM email records and lost orders placed during DNS propagation.
They needed a controlled engineering migration with verified backups and measurable before-and-after speed gains.
Our solution:
Staged server cloning, Redis object caching, and delta cutover
We followed our five-step hosting and performance runbook:
- Nginx, PHP 8.x & Redis stack: Configured dedicated PHP-FPM worker pools, Redis object caching for WooCommerce/Laravel queries, and Cloudflare edge caching.
- Delta database sync at cutover: Lowered DNS TTL 48 hours ahead, synced media files in advance, and ran a 3-minute final database delta sync during off-peak hours.
- Automated off-site backups & 24/7 monitors: Configured daily encrypted snapshots to separate object storage and 1-minute uptime and SSL expiry checks.
Before
- 1.8s to 3.2s server Time to First Byte (TTFB) on shared cPanel hosting.
- Failing mobile Core Web Vitals (LCP over 4.2s) due to unoptimized PNG/JPG assets.
- Backups stored on the same server disk with no tested restore procedure.
- Frequent 502/504 timeouts during WooCommerce traffic spikes.
- Outdated PHP 7.4 runtime and missing HTTP security headers.
After
- Sub-300ms server TTFB backed by Nginx FastCGI and Redis object caching.
- Passing Core Web Vitals on mobile and desktop after WebP and script cleanup.
- Automated daily off-site backups verified with a full staging restore drill.
- Stable checkout performance under paid ad traffic spikes.
- Upgraded PHP 8.x runtime with HSTS, CSP, and TLS 1.3 security headers.
Technologies used
Built with tools the client team can maintain and expand:
- Server Infrastructure: Ubuntu Linux, Nginx, PHP-FPM 8.x, MySQL 8, Redis Object Cache
- Security & Backups: Cloudflare WAF/CDN, Let’s Encrypt TLS 1.3, Off-Site Encrypted Snapshots
- Performance Tuning: WebP Asset Conversion, Font Preloading, Slow Query Indexing
Client feedback:
Managing Director
B2B & Retail Client
“Softlimes moved our store and portal over the weekend without a single lost order or interrupted email, and our admin panel feels twice as fast.”
Outcome:
All sites cut over with zero downtime, zero lost orders, intact business email, and passing Core Web Vitals backed by daily off-site backups.