How to use port monitoring
How to watch several devices at once, what the colours mean, and how the window tells you when the numbers have stopped moving.
Start monitoring#
- Tick the devices you want to watch in the main device list.
- Make sure the protocol at the top is Winbox 8291.
- Choose Start monitoring.
Each device gets a card. Monitoring runs only while the window is open — this is not a service that sits in the background, and once you close it, it stops collecting anything.
What a card shows#
- Every port — whether the link is up, what speed it settled on, how much is flowing right now, and what
share of the available bandwidth that is. On hardware that can do it, whether the port is supplying power. Ports are grouped by their factory names.
- The device as a whole — how many connections, CPU, memory, storage, and how long it has gone without a
restart.
- Charts — traffic per port, and CPU / memory / storage while the window has been open.
It re-reads every 3 seconds by default. You can change that at the top of the window, along with how many cards appear per row.
Reading the colours#
Colour carries the speed the port settled on, so a port that quietly dropped a tier is visible without reading any numbers.
| Colour | Meaning |
|---|---|
| Green | Up at 1G or faster |
| Teal | Up at 100M |
| Purple | Up at 10M |
| Amber | Errors or dropped packets on this port |
A port that should be at a gigabit sitting on purple is the classic bad-cable signature — and much easier to catch as a colour than as a number buried in a wall of text.
When the numbers stop moving, it says so#
A view that has quietly lost contact but keeps showing its last successful reading is far more dangerous than one that admits it cannot reach the device.
- While readings are arriving, the card just shows the current values, with no extra line.
- When a device stops answering, or a cycle brings back nothing new, the card starts showing how old its data
is — "12 seconds ago", and climbing — until fresh readings come back.
- A device that has never returned anything says "no data yet" rather than showing "0 ports", which would
read as though the device genuinely had none.
If one device stops answering it also backs off and retries more slowly, then leaves it alone after a while, so it does not hold up the refresh for everything else.
The event record#
Connects, drops, ports that keep flapping, speed changes and changes in power supply are all recorded as events, gathered into a timeline you can open per device. When someone asks "when did it go down?", the answer is in there.
Details and health check straight from a card#
Each card's title bar has two buttons:
- Details opens the full view for that device.
- Health takes that one device straight into a health check, carrying over the
kind of connection and the credentials from the main window. It opens the health check window, but it does not start checking on its own — you still confirm which groups to run.
Cutting power to a port and bringing it back#
This is the one change the monitoring window can make, and it is deliberately awkward to reach.
(Some devices supply power over the network cable to cameras, access points and the like, so this amounts to restarting whatever is plugged in.)
- Tick Allow changes (PoE power cycle) at the top of the window. Until you do, this view only looks.
- Open the port you mean.
- Trigger the power cycle and confirm it.
Next#
- Network topology — what is actually on the other end of those ports.
- Health check — a broader look, also reading only.