We delivered the rescue across four phases over six months — November 2025 to May 2026 launch — following the same framework we presented at Magento Meet Ukraine 2025. The work followed the rescue methodology by design, not by accident.

Phase 1 — Stabilization & Audit (November 2025 — December 2025)

Stage 1 of the rescue framework is "stop the bleeding." For Artizo, that meant working backward from the visible symptoms (cart bugs, slow performance, security flags) to the root causes hidden inside a 2.3.5 platform with 70+ inherited custom modules.

Work in this phase:

• Full technical audit — Magento version and patch level, server config, log analysis, indexer status, cron health, full module inventory across the 70+ custom modules from Brander

• Module classification — each custom module categorized by function (catalog, checkout, B2B/B2C, integrations, admin tooling) and by upgrade compatibility risk

• Stabilization checklist — production mode confirmed, JS/CSS merging and minification enabled, Redis configured for sessions and backend cache, image conversion to WebP, lazy loading enabled

• Security baseline — 2FA enforced for all admins, automated backups configured and verified, Content Security Policy baseline applied

• Cart bug triage — immediate fixes to the checkout-flow bugs that were directly impacting revenue, while the larger upgrade was scoped

After Phase 1, the store stopped degrading and started recovering. The cart bugs that had been costing conversion were addressed. The platform had a clean audit baseline. We had a clear picture of what the 2.3.5 → 2.4.8 staircase would actually require.

Phase 2 — Staircase Upgrade & PHP 8 Migration (January 2026 — February 2026)

The Magento 2.3.5 → 2.4.8 upgrade ran through Adobe's recommended upgrade path — not a direct jump, which isn't supported across five minor versions. The staircase:

2.3.5 → 2.3.7-p3 (final patch of 2.3 series, mandatory pre-2.4 step)

2.3.7 → 2.4.0 → 2.4.3 → 2.4.6 → 2.4.8 (sequential through 2.4.x major versions, with PHP and database compatibility checks at each step)

Critical technical transitions handled in this phase:

• PHP 7.4 → PHP 8.x — every one of the 70+ custom modules from Brander validated for PHP 8 compatibility. Code patterns that worked in PHP 7.4 (deprecated function signatures, type juggling edge cases, removed extensions) flagged and fixed

• MariaDB / MySQL compatibility — database schema validation through each upgrade step

• Composer dependency resolution — third-party module compatibility validated at each minor version; modules that couldn't survive the upgrade replaced or refactored

• Adobe Stock support, content security, and platform API changes addressed per Adobe's official upgrade guides

Each staircase step ran first on staging, with full regression testing against production data before proceeding to the next step. The 70+ custom modules were the highest-risk asset in the upgrade — Brander had built them well, but no team writes module code anticipating five major Magento version jumps. The validation work at each step was what made the upgrade ship clean.

Phase 3 — Search Migration, Integrations & Performance (March 2026 — April 2026)

With the platform now on 2.4.8 and PHP 8.x, Phase 3 covered the transitions and optimizations that 2.4.8 specifically requires — plus the performance work to bring Core Web Vitals into the green.

ElasticSearch → OpenSearch migration. As of Magento 2.4.8 (April 2025), Elasticsearch is no longer supported — the platform requires OpenSearch 2.19+. For Artizo's 10,000-SKU catalog, this wasn't an optional improvement; it was a hard platform requirement. We migrated the search infrastructure, rebuilt the catalog index, and validated faceted filtering across all category trees.

Integration verification & customization preservation. Artizo's operation depends on multiple integrations with admin-side customizations — CRM connections, ERP data exchange, custom admin workflows, and operational adapters built across the 70+ Brander modules. Every integration was tested through the upgraded platform, with the few that needed refactoring updated to maintain identical behavior on the new stack.

Performance optimization stack:

• WebP image conversion + lazy loading across the 10,000-SKU catalog

• Critical CSS extracted and inlined for above-the-fold rendering

• Varnish Full Page Cache tuned for high-traffic category and product pages

• Redis for session and cache backends, properly configured (a frequent miss on inherited platforms)

• Cloudflare CDN + WAF in front of the origin for global asset delivery and basic attack surface protection

• Database optimization — indexes added where queries were slow, fragmented tables defragmented, cron_schedule cleanup for the bloat typical of long-running stores

Code-level optimization — the same 9 code patterns we documented in our Magento 2 code optimization patterns blog post (repeated save() calls, fetchAll vs fetchCol, transaction handling for bulk operations) audited and corrected where they appeared across the inherited custom modules.

Phase 4 — Launch, Monitoring & Continuous Support (May 2026 — Present)

The upgraded Artizo platform shipped in May 2026. Post-launch work continues as an active partnership:

• Real-time error tracking via Sentry with deployment pipeline integration

• Uptime monitoring with alerts routed to the team channel that actually gets read

• Tested backup workflow — backups restored to staging on a monthly cycle, because an untested backup is just a hopeful guess

• Quarterly audit schedule — Magento moves quickly, and a store tuned in May 2026 will need adjustment in November

• Continuous development — feature additions, performance refinements, and the operational support a 10,000-SKU active platform needs

This is the rescue methodology working end-to-end: keep what was built well (the Brander foundation), fix what failed over time (the 2.3.5 platform), and stay engaged so the rescue doesn't have to happen twice.