Magento is powerful, but it is also demanding. PHP, MariaDB, OpenSearch, Redis, Varnish, queues, indexes, crons… When all of that is well-tuned, the store flies. When it is not, your TTFB, your rankings, and above all your conversion rate pay the price.
I handle that layer: the server, the operating system, and all the services Magento runs on. Magento itself — code, modules, patches, and application configuration — is handled by the agency or your development team. My job is to make sure the environment where the store lives is fast, secure, monitored, and available.
I work mainly with agencies and developers who need a systems specialist for their Magento projects without adding one to the payroll, and also directly with e-commerce managers who want a store that holds up. In the background, as an extension of your team.
Where I start: the problems I usually find
Most Magento projects arrive with one of these symptoms:
- The store is slow. High TTFB, product pages that drag, a heavy admin panel. Usually it is not the code — it is an undersized or uncached stack.
- It goes down exactly when it sells the most. Black Friday, sales, a campaign that works too well… and the server gives up at the worst possible moment.
- A migration or a new server. Hosting change, server consolidation, or a move to a better-sized environment for Magento, without losing sales or SEO along the way. (Deploying the application itself is coordinated by your team; I prepare and migrate the infrastructure.)
- Security concerns at the system level. Unsecured server, admin panel accessible from any IP, no WAF, no clear backup plan. In e-commerce, that is money and customer data at risk.
If any of those sound familiar, that is where I come in.
How I solve it
I do not sell magic recipes or generic “optimisations”. The approach is always the same: measure, dimension, and tune the specific stack of your store, not the one from a manual.
I start from real data (server metrics, service profiling, traffic logs from past campaigns) to find where the actual bottleneck is before touching anything. From there I tune each service in the stack — web server, PHP, database, cache, search — for the real load pattern of your store, leave everything documented and reproducible, and stay available for peaks. The idea is to have the infrastructure ready before the campaign, not fighting fires during it. The application runs on top; I make sure it has solid ground beneath its feet.
What is included
This is what I contribute at the system level in a Magento project. You do not need to take all of it: it adapts to what the project needs.
Performance and stack tuning
- Tuning of PHP-FPM and OPcache for the store’s real load pattern (pools, workers, memory).
- Provisioning and tuning of the Redis service Magento uses for sessions and cache (instances, memory, persistence).
- Varnish deployed and configured as a cache in front of the store, coordinating with your team on Full Page Cache integration.
- Tuning of MariaDB/MySQL (buffer pool, connections, slow queries) according to the catalogue and traffic.
- Provisioning and sizing of the OpenSearch / Elasticsearch service that powers search.
- Core services properly configured (RabbitMQ, the cron daemon) so that the Magento tasks that depend on them have somewhere solid to run without crashing the server.
Campaign and peak preparation
- Pre-event load tests to know the server’s real ceiling before the event.
- Resource sizing (vertical and, where applicable, horizontal with load balancing) to handle the peak.
- Contingency plan and reinforced availability (on-call) during the critical window.
Security and hardening
- Server hardening: firewall, fail2ban, IP-restricted admin panel access, permissions, and web server-level headers.
- System security updates: OS, kernel, and stack services kept up to date. (Magento application patches are your team’s responsibility; I keep what is underneath secure.)
- WAF in front of the store (Cloudflare, ModSecurity, or others) and bot and scraping mitigation.
- Best practices aligned with PCI-DSS compliance on the infrastructure side.
- Management of TLS certificates and automatic renewals.
Infrastructure migrations
- Hosting or server migrations with no perceptible downtime and no SEO loss.
- Provisioning of equivalent staging and production environments, so your team deploys and tests on a solid base.
- Data and service migration at the system level, with a clear rollback plan if something goes wrong.
Backups and monitoring
- Encrypted, offsite, tested backups (3-2-1 rule): a backup that has never been restored does not count.
- Continuous monitoring of server, services, and availability — so you find out before your client does.