Global Website Network Delivers Long-Term Value  

The Situation

The client’s content network was started by a very small team of content producers and built-up site by site. The expectation was that the publishing workflow and the operational maintenance of all the websites would be as easy and frictionless as running as early WordPress blogs. While very friendly to the publishers, the WordPress platform lagged behind on industry best practices for security and development workflows. The websites’ cybersecurity focus made them a prime target for hackers looking for ways to sharpen their skills or have some laughs. Meanwhile, the client’s content marketing strategy ambitions extended to 14 languages around the world with a fast and clean experience for visitors.

Our Approach

With the client firmly in charge of the content production efforts, we set out to support them with the technical aspects of running the websites in a manner that maximized productivity and sites security. After considering various technologies options carefully, we determined that evolving the current crop of WordPress websites was both possible and the best way to build on top of their existing investment without interruptions. Convertiv identified gaps in their setup related to scalability, security, and performance and then set off to implement the following:

• A multistage development workflow was implemented, enabling the development team to test and evolve large chunks of the codebase without interruption to the live website. A virtualized local environment was provided to all team members, along with tooling that allowed them to build and test locally on their machines, stage for client review, and ultimately deploy their work to the live production site.
• Continuous deployment was introduced to support the ever-growing speed of improvements and changes to the website. It allows the development team to deploy changes to the website multiple times per day with confidence.
• A regular update cadence was introduced to keep up with all the updates to third party code dependencies. This was streamlined across the environments by introducing code-based package management in lieu of the normal WordPress plugin workflow. This allowed the process to be as efficient and reproducible from development to production.
• Critical third-party code vulnerabilities were identified and patched as soon as possible via our internal system. It notified us of any new vulnerabilities for any of the plugins or other code used on the websites, and we quickly initiated remediation efforts when needed.
• A SIEM and performance dashboard was built using the ELK stack. By collecting various application events (e.g., logins or uploads) and ingesting various server logs, our operations team had eyes on everything happening with the website from a single pane of glass. This allowed us to identify security events or technical issues affecting site availability or performance quickly and react promptly.
• WordPress itself has a long technical legacy, and it was not created with the scalability of the cloud in mind. Working with our partner vendors, we deployed WordPress successfully on a new horizontally scalable and fault-tolerant architecture running on AWS. This ensured stable performance and availability of the content network.

The Result

• Scaled to approximately 90 websites in the network
• Internationalized sites for approximately 20 countries
• Operated incident-free for seven years without
• Continuing feature and update rollouts multiple times per day