When a 9.2/10 PHP security vulnerability hits your business

Written by Michael Staněk

Published: October 27, 2026

While writing this article, the URL generator turned the “PHP 9.2/10” headline into something that looked remarkably like Beverly Hills 90210. I couldn't help but smile. The ’90s bring back some good memories. This incident, however, came with considerably more drama.

On October 21, 2026, a critical PHP security incident sent administrators, hosting providers, and website owners scrambling to protect their systems.

At RedmineX, it meant temporarily taking parts of our website infrastructure offline, interrupting customer payments, and spending several days coordinating recovery.

The experience reinforced two things I consider essential to running a technology business: sound risk management and having the right people responsible for critical infrastructure.

When you decide to shut down your payments

I first learned about the incident on October 22 through a notification from one of our hosting providers. It sounded more serious than the usual security advisory, but operations continued normally for a few hours. Later during the day, I learned about the severity of the situation: 9.2 out of 10.

Once I understood the potential implications, I spent the next couple of hours evaluating our exposure and the available options.

Our customer portal handles payments, subscriptions, product downloads, and account management. Taking it offline would create immediate inconvenience for our customers and interrupt revenue.

But security had to take priority over availability.
Better safe than sorry.

We decided to temporarily disable the potentially volnurable applications. Our main public website remained operational, while the customer portal and payment functionality were taken offline. Another website was replaced with a static HTML version.

Before shutting anything down, I preserved the necessary backups and ensured the applications could be recovered without unnecessary data loss.

It wasn't a pleasant decision, but it was a deliberate one. Sometimes protecting a business means temporarily interrupting it. Apparently.

Critical risk management without AI

Then came a series of decisions.

  • Which services were affected by the shutdown?

  • What would customers experience?

  • How should we communicate the interruption?

  • Which links will now stop working?

  • What about subscriptions and recurring payments?

  • What needed immediate attention, and what could wait?

The challenge was evaluating all of the questions together, understanding their dependencies and business impact.

I rely heavily on AI for analysis, troubleshooting, challenging assumptions, and drafting communications. But in this situation, I instinctively knew it wouldn't be particularly helpful. The relevant context was spread across years of business experience, operational knowledge, financial considerations, and technical dependencies.

Out of curiosity, I tried asking AI for help anyway. The response was so disconnected from what I actually needed that I didn't even finish reading it.

What surprised me was how reassuring that experience felt. Despite relying so heavily on AI these days, I hadn't lost the habit of working through complex problems without AI — which was something that bothered me in the back of my head.

Fortunately, not alone

Once the initial decisions were made, the focus shifted to execution.

We needed to communicate the interruption, ensure customers could still find essential information, preserve access to support & downloads, and prepare the affected systems for restoration once the server PHP was updated.

There were dozens of small but interconnected tasks: checking endpoints, updating website notices, reviewing links to customer services, coordinating technical changes, and verifying what remained operational.

Fortunately, we weren't handling the technical side alone.

Our infrastructure specialists, developers, and hosting providers were responsible for their respective areas. That allowed us to divide the work, make decisions quickly, and move toward recovery without improvising an infrastructure strategy in the middle of an incident.

Over the weekend, we prepared the affected applications for the updated PHP environment and coordinated the necessary changes.

On Monday, October 26, the hosting-side fix became available, allowing us to bring the customer portal and payments back online.

Why managed infrastructure matters

Perhaps the most important takeaway wasn't about PHP at all. It was about the value of having experienced professionals responsible for critical infrastructure.

At RedmineX, infrastructure is part of our core business. We operate and maintain Redmine environments for customers who depend on them every day. We know what goes into keeping these systems running: security maintenance, updates, monitoring, backups, compatibility testing, incident response, and continuous operational oversight.

This incident was a reminder of why those responsibilities matter.

Our own website infrastructure involves multiple technologies and service providers, each managed by people with the relevant expertise.

When the security warning arrived, we had specialists to consult, established operational responsibilities, and people capable of executing the necessary changes.

What this means for Redmine users

At RedmineX, we provide managed hosting and maintenance for Redmine, which is often a business-critical project management or ticketing application.

It's tempting to compare the price of managed hosting with the cost of renting a server and conclude that self-hosting is cheaper.

But the server is only one part of the equation. Security updates, compatibility testing, monitoring, backups, incident response, and qualified personnel all have a cost. So does downtime.

Self-hosting can absolutely make sense for organizations with capable internal infrastructure teams. For everyone else, outsourcing those responsibilities to a competent managed service provider is often a much safer and more economical choice.

The question isn't whether you manage infrastructure internally or externally. It's whether someone is clearly responsible for keeping it secure, updated, and recoverable.

And when something goes wrong, you need more than a support ticket and a promise to investigate. You need people who understand the technology, know the environment, and can take responsibility for getting things working again.

That's the value we aim to provide our customers.

This was a preview

I don't expect the October 21 incident to be the last major security challenge the industry faces. What matters is how prepared an organization is to respond. This experience reinforced two convictions.

First, critical risk management requires human judgment. AI is an excellent tool, but business context, prioritization, and accountability cannot simply be outsourced to a language model just yet.

Second, professional infrastructure management is not merely about keeping servers online. It's about ensuring that the expertise, processes, and responsibilities are in place when something unexpected happens.

Security isn't the absence of incidents. It's the ability to recognize risks, make the right decisions, and respond effectively.

And unlike Beverly Hills 90210, this is one series I'd prefer not to see renewed for another season.

Stop babysitting your Redmine.

Share this if you liked it!

Take a look at our Youtube Channel.

Trusted By Effective Teams Worldwide

Try All RedmineX Plugins

entire solution & plugins  |  totally FREE

  • All plugins
  • RedmineX Cloud
  • On-premise managed Redmine
  • UX Upgrade
  • Gantt chart
  • Resources
  • Agile MyPage
  • Project templates
  • Statistics
  • Inline edit
  • Timesheet
  • Microsoft Auth
  • Secure Async Export
Source? *

How Does The Demo Work?

  1. Login credentials
    After submitting the form, you will be redirected to a page with login credentials.

  2. One demo to rule them all
    The demo site is shared and the database is being reset every full hour. You can do anything, but the data you enter will eventually be reset. You can qualify for a dedicated demo if you‘re interested in keeping the data, or testing more thoroughly.

  3. Ask anything
    Michael is ready to help you with any request so you can squeeze maximum from your Redmine.