Introduction
In most medium-sized infrastructures, we often talk about redundancy for the most visible elements: critical servers, main internet links, backups…
However, when we look closer, we very frequently find single points of failure that are not always identified as such.
These fragilities generally do not cause problems on a daily basis. They only become critical when an incident occurs… and it’s often at that moment that we realize we didn’t have a real plan B.
Why Do These Fragility Points Go Unnoticed?
Most SME infrastructures were not designed all at once. They have grown gradually.
At each step, a pragmatic solution that « works for now » is added.
ℹ️ Info: The result is often an architecture that works well… as long as no important element fails.
Here are the situations I encounter most often in the field.
1. The Single (or Almost Single) VPN Tunnel
This is probably the most common case.
A company has multiple sites interconnected via Site-to-Site VPN tunnels. On paper, everything is redundant… but in reality, a single tunnel often carries most of the critical traffic.
⚠️ Warning: When this tunnel goes down, the consequences can be immediate: inability to access central applications, disruption of business flows, and sometimes even partial business shutdown.
2. The Central Equipment That Does "Too Much"
Many SMEs have one (or two) firewalls/routers that concentrate a large number of functions: routing, filtering, VPN, proxy, monitoring, authentication…
This equipment then becomes a major single point of failure.
🔴 Important: Its failure or saturation simultaneously impacts multiple critical services.
3. Poorly Redundant Cross-Cutting Services
Some services are used by the entire infrastructure but do not always have the same level of redundancy as business systems.
We are thinking in particular of: Active Directory, internal DNS, monitoring, centralized backups.
⚠️ Caution: When one of these services becomes unavailable, the consequences can be widespread and sometimes surprising.
How to Objectively Identify Your Fragility Points?
The best way to proceed is to ask yourself a simple question:
If this element fails tomorrow morning, what concretely happens to the business?
💡 Tip: Ask yourself these three questions:
- Is there a real operational workaround solution?
- Has this solution been tested recently?
- Who would be impacted, and for how long?
How to Realistically Address These Fragilities?
In most SMEs, there is no question of aiming for « enterprise » resilience. The goal is to identify the most important risks and provide proportionate responses.
📝 Note: The important thing is not to achieve perfection, but to significantly reduce the risk that « everything stops at the same time ».