← All posts Deutsch

Why One Device Can Have Several IPv6 Addresses

An IPv6 address list can easily look like a device list. A router, network scanner, or operating-system settings page may show two, three, or more IPv6 addresses next to one computer. When the endings differ, it's easy to suspect that a second device has appeared, an old record has stuck around, or somebody has joined the network.

Several IPv6 addresses on one network interface are usually normal. IPv6 gives an interface a link-local address. The network may also give it one or more addresses that work beyond the local link, and privacy features can add temporary addresses. Address lifetimes can overlap while a replacement is being prepared. So the list can change while the device stays connected. Before counting each address as a separate device, check which interface it belongs to, its scope, and when it was observed.

If you're starting with an entry whose owner you don't know, begin with How to Investigate an Unknown Device on Your Home Network. The question here is narrower: why one known device can have several IPv6 addresses without those addresses representing several physical devices.

The interface behind the address

An IPv6 address belongs to an interface. A laptop can have Ethernet, Wi-Fi, a VPN adapter, a virtual-machine bridge, or a container network at the same time, each with its own addresses. Even without privacy addressing, a list that includes all interfaces isn't an inventory of people or hardware.

Start by recording where you saw the address. A router's client page may describe a Wi-Fi association, while a computer's network settings describe its local interfaces. A scanner may report an address that answered from the subnet where the scan ran. Each source can be useful. They don't necessarily show the same addresses, though, or update at the same time.

Next, find the relevant interface. If a computer is connected by Wi-Fi, compare the address with that computer's Wi-Fi interface, not an inactive Ethernet port or a VPN interface. For a NAS, use the port and VLAN it actually uses. That keeps the comparison on one network path and stops unrelated addresses from getting folded into a device history that never happened.

Don't use the address ending to decide who owns it. An interface identifier may come from a MAC address, be generated as a stable opaque value, or be generated temporarily. A similar suffix is worth checking, but it doesn't prove that two entries belong to the same device. Different suffixes don't prove that they belong to different devices, either.

Common IPv6 address types

Every IPv6-enabled interface normally has a link-local address, usually beginning with fe80:. This address works only on the local network link. Routers use link-local addressing to exchange local IPv6 information, and an operating system can use it even when the network has no usable public IPv6 connectivity. It isn't an Internet address. Don't copy it into a browser URL without the interface context required by the operating system.

You may also see a global unicast address (GUA), commonly beginning with a prefix delegated to the home connection. The network provider and router choose the exact prefix. Having that address doesn't mean every inbound connection from the Internet is allowed. The router firewall and device firewall still decide what can reach a service.

Some networks use unique-local addresses (ULAs), often beginning with fd. These have global scope but are intended for internal routing, not public Internet routing. Scope does not establish public Internet reachability. A device can have a unique-local address alongside a link-local address and a global unicast address. Those are different ways to reach the same interface, not three machines.

Write down the full address, its prefix length if displayed, the interface, and the time. Keep the full addresses out of support posts: a prefix can identify a household or organization more precisely than people expect. A redacted prefix and a description of the scope are often enough for troubleshooting.

Temporary addresses on a known device

IPv6 privacy extensions can create temporary addresses for outgoing connections. The goal is to make long-term tracking through a device’s IPv6 addresses harder, including while it stays on the same network. A current operating system may keep a stable address for receiving local connections and use a temporary address for some outbound connections. Both can appear in the address list at once.

Temporary addresses have preferred and valid lifetimes. Before an existing temporary address stops being preferred for new connections, the system can create another. For a while, you may see both the older temporary address and the newer one. That overlap is expected. It doesn't show that the device restarted or that another device impersonated it.

Operating systems use different names and controls for these features. Don't turn privacy addressing off just to shorten a router's list. Doing so can change how the device appears on that network and make later comparisons harder. If you administer the device and need to identify it, first check the address it shows for the specific Wi-Fi or Ethernet network. Record the setting and any planned change before you change it.

Be just as careful when interpreting router entries. A broadband reconnection can change the IPv6 prefix delegated to a home router. Devices may then configure replacement addresses while earlier entries remain in a router or scanner cache. All that tells you is that the observed address changed. A provider change, a router reboot, a device reboot, or ordinary address-lifetime behavior could explain it, but each possible cause needs its own evidence.

A small, repeatable comparison

Choose one known device that you administer. Check its own network settings and one other view, such as the router's active-client page, and capture both within a short interval. Don't disconnect cameras, alarms, storage, or other important equipment to force a test.

For each observation, note:

  1. The date, time, and timezone shown by each system.
  2. The interface and network name or VLAN, if known.
  3. Each IPv6 address with its type: link-local, global unicast (GUA), or unique-local (ULA).
  4. Whether the operating system labels an address temporary, deprecated, preferred, or stable.
  5. The device's MAC address only in a private local record, if you need it to compare a router entry.

Repeat the observation later without changing the network. If a temporary address has changed but the interface, known device, and other evidence are the same, record exactly that. It explains why an unfamiliar IPv6 entry has appeared in the list. If the router shows two simultaneous active client records with different hardware identities, investigate those records separately. Don't combine or dismiss them based on IPv6 suffixes.

A broad scan shouldn't be your only way to settle the question. A host can decline to answer a particular probe, or a firewall can filter it. A missing scan result doesn't mean the address is stale. A response tells you that an address answered from the observed network path at that time. It doesn't identify the person or physical object behind it.

Service checks by address family

An address in a list doesn't tell you which services are available on it. A NAS might advertise a file service on IPv4 and IPv6, on only one, or on neither to a particular client because of a firewall rule. A phone may have several IPv6 addresses and accept no incoming connections at all. Record address assignment separately from reachability and service availability.

When a service matters, test only a service you're authorized to use and record which address family the client used. For example, a managed NAS's own status page may show an uptime while a file-service connection works over IPv4 but fails over IPv6. That tells you something about a particular route and service. It doesn't make every IPv6 address invalid.

Network boundaries matter, too. Guest Wi-Fi, client isolation, VLANs, and routed subnets can leave an address visible to a router but unreachable from a laptop. Before changing a router setting, compare the network the client was on when it made the observation. An IPv6 address that's appropriate on one VLAN may be deliberately unreachable from another.

Conclusions supported by the evidence

A useful note can be short: “At 14:10 local time, the Wi-Fi interface of this known laptop had one link-local address, one global stable address, and one temporary global address. The router listed the same hardware identity.” It records what was seen without inventing a cause.

Stronger claims need stronger evidence. “A new temporary IPv6 address proves the laptop was replaced” isn't supported. Neither is “Two addresses prove the device is exposed to the Internet”. Before deciding whether a device is exposed, compare the router's firewall policy, the device firewall, and an authorized service test.

For a local record, DeviceShelf is one possible way to compare observed addresses and device details over time. Each observation is a lead to check against the router and the device itself. It isn't proof of device ownership, availability, or a security event. Checking it this way keeps normal IPv6 address rotation from turning into an unnecessary incident.

Sources checked

Network tips and product updates

The occasional guide, new features and release news, straight to your inbox. No spam, unsubscribe anytime.