Introduction
Today I want to share a real case study from my own infrastructure. No buzzwords — just a practical story about how ambiguous communication led to repeated internet outages every five minutes.
I pay for business fiber connectivity, even though I use it at home (and at my parents’ house). Since these are not production business sites, I do not pay for fallback connectivity.
This detail matters.
The SMS That I Ignored
One day my ISP sent me the following SMS:
Dear customer,
We are contacting you regarding a replacement of a power adapter to your router. This replacement is important and makes sure your internet connectivity stays stable.
You can ask for a replacement, free of charge at .
Most people probably either comply immediately or ignore such messages entirely.
I ignored it — but for different reasons.
- I operate three separate sites. Two were referenced in the message. However, I could not map cryptic IDs like
zkhradng375to physical addresses. Those identifiers make sense internally to the ISP. They are meaningless to customers. - I had zero outages at the time. All sites were stable.
- The explanation felt oversimplified. “It makes your internet stable.” How exactly? It is a power adapter. If it outputs 12V and the router LEDs blink, everything should work.
Over the next months, I received two or three similar reminders.
I ignored all of them.
The First Symptoms
Eventually, my parents began experiencing short outages. Streaming services would pause and reconnect every few minutes.
Intermittent outages are worse than complete downtime. They create uncertainty.
I started diagnostics.
The situation was complicated by the fact that I run my own equipment behind the ISP router. A faulty Ethernet port on my own router would not have been unprecedented. Consumer hardware degrades.
On inspection, I observed a storm of PPPoE messages.
The router would:
- Dial in
- Stay connected briefly
- Disconnect randomly
- Repeat
This pattern narrowed the problem to three possibilities:
- ISP-side issue
- My router failing
- ISP router failing
The demarcation point between my router and theirs defines responsibility. The ISP guarantees service up to their router.
Before calling support, I wanted evidence.
Hypothesis Elimination
Step 1 — Wait
If the issue was on the ISP side, it would likely affect more households. I waited two to three days.
The outages persisted.
No public outage reports.
Hypothesis 1 eliminated.
Step 2 — Replace My Router
I already had a router upgrade planned. I reprioritized and replaced my MikroTik with a new OPNsense deployment.
The result?
The issue became worse.
This narrowed the failure to a flapping Ethernet link between my router and the ISP router.
Now the demarcation logic was clear.
The Support Call
I contacted ISP support and presented:
- Observed PPPoE flapping
- Router replacement already performed
- Issue reproducible
The technician asked a simple question:
“Did you replace the power adapter?”
This time, the context was technical, not marketing.
I asked the key question:
How can a power adapter cause Ethernet instability when fiber link indicators appear stable?
The explanation:
Older adapters can degrade. Voltage may fluctuate microscopically. The router appears powered, but Ethernet PHY layers are sensitive to instability. Micro voltage dips can cause link renegotiation or brief resets without fully powering down the device.
This was plausible.
I replaced the adapter the next day.
The connection stabilized immediately.
Technical Interpretation
The failure mode likely involved:
- Aging power supply
- Slight voltage instability under load
- Ethernet interface reset behavior
- PPPoE reconnection loop
The router never appeared “off.”
But networking is unforgiving at the physical layer.
Power integrity matters.
Lessons Learned
1. Communication Layering Matters
If you are a service provider:
Provide two layers of explanation:
- Layman summary
- Technical reasoning
Without the technical explanation, technically literate users may dismiss your message as marketing noise.
In my case, the oversimplified wording reduced credibility.
A single sentence explaining voltage degradation and Ethernet sensitivity would have prevented months of dismissal.
2. Technical Confidence Can Become Blindness
If you are technically experienced:
Do not ignore signals simply because they appear oversimplified.
One call to support could have saved:
- Several days of investigation
- Router replacement time
- Diagnostic overhead
Expertise should reduce friction, not create it.
3. Systems Fail at Interfaces
The outage did not originate in fiber.
It originated at the interface between power delivery and network signaling.
Most real-world failures occur at boundaries:
- Hardware ↔ OS
- Power ↔ signal
- Provider ↔ customer
- Abstraction ↔ implementation
The ambiguity was not technical.
It was communicational.
Current State
One remaining site still uses the old adapter. It is located inside a locked server room, so I postponed replacement.
However, I now know the symptoms:
If PPPoE begins flapping, I can resolve the issue within hours.
Clarity converts outages into procedures.
Closing Reflection
As a systems architect, I focus on calibrated clarity.
Granularity matters.
Too little detail alienates technical audiences.
Too much detail overwhelms non-technical ones.
Ambiguity is not neutral.
It compounds into friction.
In this case, ambiguity caused months of delay and multiple outages — even though the solution was free and available from the beginning.
Communication is part of infrastructure.
When designed poorly, it fails like any other subsystem.
