Understanding 802.1Q Trunks
Today's lesson is Understanding 802.1Q Trunks. The answer depends on more than the word trunk. You need to separate three questions. Is the interface operating as a trunk?
How is native VLAN traffic expected to be handled? Which VLANs are actually permitted across the link? Those questions describe related parts of the same connection, but they don’t have identical answers. A link can be operational as a trunk while still excluding a particular VLAN.
A useful way to organize this episode is around three checkpoints. A trunk carries traffic for multiple VLANs. Allowed VLANs determine which VLANs may cross. Native VLAN expectations must agree.
Together, those checkpoints separate the durable concept from the exact notation you will inspect later. First, define what should be true. Think about what this does and doesn’t change. The trunk provides a path that can carry multiple VLANs.
It doesn’t merge those VLANs into one broadcast domain, and it doesn’t turn Layer Two switching into routing. VLAN ten remains distinct from VLAN twenty. The trunk extends the relevant Layer Two paths between switches. Communication between different VLANs still requires Layer Three forwarding, while communication within the same VLAN can use the trunk when that VLAN is eligible to cross it.
That distinction gives you a useful mental model. Trunk mode describes the role of the connection. The native VLAN describes a tagging expectation for native traffic. The allowed list places a boundary around the VLANs the trunk may carry.
When you configure or interpret a trunk, you need all three parts. Looking at only one can leave you with a technically true statement that still doesn’t explain the observed forwarding result. Notice that the native setting and allowed list answer different questions. Choosing ninety-nine as native establishes the intended treatment of native trunk traffic.
It doesn’t mean ninety-nine is the only VLAN on the trunk. Likewise, allowing ten and twenty means those VLANs are eligible to cross. It doesn’t make either one native. If you collapse those ideas together, you can create a configuration that sounds reasonable while failing to represent the actual requirement.
Configuration begins with the intended interface role. On each switch, the interswitch interface must be placed in trunk mode rather than treated as an ordinary access interface. You then establish the required native VLAN and define the allowed VLAN set. The exact configuration is an input to the switch.
It expresses what you want the interface to do, but a command being accepted alone does not establish that the resulting operational state matches your design. Now consider VLAN thirty. It exists on both switches, and devices have been assigned to it on both sides. Is that enough for VLAN thirty traffic to cross this trunk?
No. A VLAN that exists on both switches still can’t cross a trunk when it isn’t permitted by that trunk’s allowed list. Local VLAN existence and trunk eligibility are different conditions. The VLAN database can tell you that the switch recognizes VLAN thirty locally without proving that this particular interswitch connection will carry it.
That explains why employee traffic might work while guest traffic fails. If the link is operational as a trunk and VLAN ten is allowed, employee frames have an eligible path across it. If VLAN thirty is absent from the allowed list, guest frames don’t gain that same path merely because the interface is called a trunk. The successful employee test proves something about VLAN ten.
It is not enough to establish that every configured VLAN is eligible to cross. Exact visual detail related to trunk configuration, allowed-list examples, and tagged-frame diagram is preserved in the episode reference. The audio teaches the relationships and decisions. The reference supplies the characters, rows, fields, and examples needed for deliberate practice.
Then identify what would make that state visible. Suppose the requirement changes and guest traffic must now use this link. Your prediction should come before the configuration change. The trunk role can remain the same.
The native VLAN requirement can also remain ninety-nine. The relevant change is to the allowed list, which must now permit VLAN thirty. That’s a focused decision based on the missing eligibility condition. Replacing the native VLAN or recreating every local VLAN would change variables that don’t address the stated constraint.
After adding thirty to the intended allowed set, don’t stop because the device accepted the configuration. Compare the expected state with operational evidence. Is the interface actually operating as a trunk? Does the switch report the intended native VLAN?
Is VLAN thirty shown among the VLANs allowed on the trunk, and does the operational information show it among the VLANs that can currently forward there? Those observations connect the requirement to the forwarding state the switch is really using. This is where configuration text and operational state must be kept separate. Configuration text records requested settings.
Verification evidence reports what the switch is doing under current conditions. Both matter, but they answer different questions. A line in the configuration may express trunk mode, a native VLAN, or an allowed list. Verification asks whether the interface is operationally a trunk and which VLANs are actually allowed and forwarding over it.
Imagine that VLAN twenty voice traffic now fails between the switches. The first temptation may be to change several settings at once. You might replace the allowed list, change the native VLAN, and reset the interface role. That produces activity, but it weakens the diagnosis.
If traffic begins working afterward, you won’t know which change corrected the problem. You may also introduce a second difference while hiding the first. Start instead with an explicit prediction. VLAN twenty should cross only if the interswitch interface is operating as a trunk and twenty is eligible on that trunk.
The trunk’s tagging expectations should also reflect the intended native VLAN arrangement. Now collect evidence for those conditions. Don’t begin with the assumption that every configured VLAN is available just because some traffic crosses the link. This sequence keeps the evidence tied to the symptom.
First confirm the link’s operational role. Then examine the native VLAN expectation. Then inspect the VLANs actually allowed and forwarding. The sequence isn’t meant to imply that every fault has the same cause.
It prevents one successful observation, such as seeing the word trunk, from being treated as proof of every other required condition. Now test the most tempting wrong conclusion. Suppose the evidence shows that the interface is operationally trunking, the intended native VLAN is reported, and VLAN ten is forwarding. VLAN twenty, however, is missing from the allowed set.
That’s a specific difference between expected and actual state. Correct the allowed list rather than changing unrelated settings. Then repeat the same operational checks. VLAN twenty should now appear where the forwarding state indicates that it’s eligible to use the trunk.
Repeating the verification matters. The disappearance of one visible symptom isn’t as informative as a result tied to a known mechanism. You predicted that permitting twenty would change its trunk eligibility. After the correction, observed operating state should support that prediction.
You can then explain both the failure and the recovery. The trunk was capable of carrying multiple VLANs, but its allowed list excluded the affected one until the configuration was corrected. Before implementation, make several predictions. Traffic in ten should be eligible to cross.
Traffic in twenty should also be eligible. The same is true for forty. Traffic in thirty shouldn’t use this trunk, even though both switches recognize that VLAN locally. Native traffic should be associated with the configured native VLAN.
These predictions give you a standard against which to interpret the resulting state. After configuration, verification should answer the same questions in operational terms. Is the interface shown as a trunk rather than merely configured with trunk-related settings? Is the reported native VLAN ninety-nine?
Does the allowed set contain ten, twenty, and forty while excluding thirty? Among the allowed VLANs, which ones does the switch report in the relevant forwarding state? A careful answer distinguishes what is configured, what is permitted, and what is currently operational. Now imagine that thirty appears in the allowed list despite the requirement to keep it off this connection.
Communication might still appear normal for the VLANs you intended to carry. That doesn’t make the state correct. Verification isn’t limited to asking whether desired traffic works. It also checks whether the operational boundary matches the design.
In this case, the trunk permits more VLANs than required, so the allowed list should be corrected and checked again. The opposite difference is easier to notice because it causes a missing path. If forty is absent from the allowed set, the VLAN can exist on each switch while failing to cross this trunk. Adding or confirming the local VLAN alone doesn’t correct the trunk-carriage condition.
You must address the allowed list and then verify the operational result. That’s why local VLAN existence can’t be used as a substitute for trunk verification. Close the loop by verifying the result after one controlled change. Native VLAN verification deserves its own attention because it represents a tagging expectation, not a general permission switch for every VLAN.
If the requirement identifies ninety-nine as native, inspect the operational trunk information for that value. Don’t infer it from the allowed list, and don’t assume that selecting ninety-nine as native has made ten, twenty, or forty eligible. The native setting and allowed list should each be compared with their own requirement. You can organize any trunk decision into a chain.
Begin with the required traffic. Identify which VLANs need an interswitch path and which ones shouldn’t use it. Establish the intended trunk role and native VLAN. Translate the required paths into the allowed VLAN set.
Then predict the measured network behavior you expect to see. Only after that should you decide whether the state is correct or which focused change is justified. This chain is also useful when interpreting a proposed configuration. Don’t choose an answer merely because it contains familiar trunk terminology.
Ask what state that configuration is intended to create. If the requirement says VLANs ten and twenty must cross, an allowed list containing only ten doesn’t satisfy it. If the requirement names a nondefault native VLAN, the native setting must reflect that requirement independently of which other VLANs are allowed. At the same time, avoid turning configuration into memorized syntax without a model.
Syntax can express trunk mode, native VLAN selection, and allowed VLAN membership, but it can’t decide the design for you. If you don’t know which VLANs should cross, no command sequence can supply that missing requirement. If you don’t verify operational state afterward, a plausible configuration can give you confidence without proving the forwarding behavior. The original failure now has a precise explanation.
Employee traffic crossed because its VLAN had an eligible path over an operational trunk. Guest traffic failed because calling the link a trunk didn’t automatically permit the guest VLAN. The right response is to compare the requirement with trunk state, native VLAN expectations, and the allowed and forwarding VLAN information. That comparison tells you whether guest traffic is intentionally excluded or whether the allowed list needs a focused correction.
Use one final decision to test the model. Bring the model together. A trunk carries traffic for multiple VLANs. Allowed VLANs determine which VLANs may cross.
Native VLAN expectations must agree. Use the episode reference for the exact notation, then prove the result with device-reported state in a lab. Knowledge check: When one VLAN crosses a trunk and another does not, what condition should you inspect first? Answer before continuing.
Explain what you would expect, what you would observe, and how the observation would guide the next action. Thanks for listening. This lesson was produced by Bare Metal Cyber dot com.