You are staring at your router or DHCPv6 server console, with options like Rapid-Commit, managed flags, and DNS settings staring back. You will be able to stand up working IPv6 address assignment by following a step-by-step configuration that clarifies what each choice controls. All you need is your current IPv6 prefix, the interfaces involved, and administrator access to apply the changes safely.
Configure IPv6 DHCP by choosing stateful vs stateless DHCPv6, then making sure your DHCPv6 server settings and Router Advertisements (RA) don’t contradict each other. In practice, most “it’s running but nothing works” failures happen when RA tells clients to behave one way, while the DHCPv6 server is configured for the other.
If you’re setting up an IPv6 network for a home lab, office, or small business, this guide focuses on the core concepts and the exact checks that typically prevent “it’s running but nothing works” scenarios.
Clarify what “IPv6 DHCP” means in your network
You should decide up front whether clients should get addresses from DHCPv6 (stateful) or from RA (stateless), because clients key off RA signals before they ever start using DHCPv6. Then you align DHCPv6 options (like DNS) so they complement—not conflict with—RA.
– Choose the mode: stateful DHCPv6 (DHCPv6 hands out IPv6 addresses) vs stateless DHCPv6 (RA provides addressing; DHCPv6 supplies extra info like DNS).
– Confirm how your router handles IPv6 Router Advertisements (RA), because DHCPv6 clients react to RA flags and timers.
– Identify where addressing comes from: a fixed IPv6 range, a delegated prefix (often via Prefix Delegation), or an ISP-provided prefix.
To make this concrete, remember that RA is not “optional fluff”—it’s the mechanism that tells clients how to configure their addresses. In one troubleshooting case I’ve run into repeatedly (across multiple lab setups), a network was configured for *stateful* DHCPv6, but RA flags implied *stateless-only* behavior. The DHCPv6 server was issuing replies, yet clients never installed the leases because the client-side expectation was wrong.
Stateful vs stateless: which one should you pick?
If you want the simplest mental model, pick stateful DHCPv6 when you want centralized, repeatable control of address assignment. Pick stateless DHCPv6 when you’re comfortable with RA handling addressing but still want consistent DHCPv6-delivered data like DNS and NTP.
Mode comparison (scanable decision table)
| Feature | Stateful DHCPv6 | Stateless DHCPv6 |
|---|---|---|
| Who assigns addresses? | DHCPv6 server | RA (router) |
| What does DHCPv6 provide? | Addresses + options | Options only (DNS/NTP/etc.) |
| Address tracking for clients | Centralized leases | Less centralized (RA-driven) |
| Common failure mode | RA/client flags mismatch | Clients expect DHCPv6 DNS but don’t get it |
| Best for DNS consistency | Yes (everything centralized) | Yes (DNS via DHCPv6) |
| Change control | Predictable via lease lifecycle | More tied to RA update cadence |
| Operational complexity | Higher (pools/leases) | Moderate (RA + options) |
| IPv6 autoconfig dependency | Still needed (initial discovery) | Core dependency (addresses) |
| Typical lab fit | “I want control” labs | “I want minimal moving parts” labs |
| Best for when upstream prefix changes | Adjust DHCPv6 pool mapping | Update RA prefix info; DHCP options stay stable |
| Best For | Centralized addressing + options | RA addressing, DHCPv6 options |
Stateful vs stateless: the “you’ll notice this first” view
| Criteria | Stateful DHCPv6 | Stateless DHCPv6 |
|---|---|---|
| Client first signals | RA discovery + then DHCPv6 address request | RA addresses immediately; DHCPv6 for options |
| What you validate first | DHCPv6 lease/pool match | RA prefix + DHCPv6 DNS delivery |
| Connectivity symptoms when misaligned | Clients have no usable leased addresses | Clients connect but DNS fails (most common) |
| Firewall implications | Must allow DHCPv6 server responses on LAN | Still needs DHCPv6 traffic allowed for options |
| Change workflow | Pool/lease changes propagate over time | RA updates propagate; DNS stays via DHCPv6 |
| Best for repeatable “known addressing” | Strong fit | Weaker fit |
| Operational overhead | More moving parts | Less moving parts than stateful |
| Common monitoring signals | Solicit/Advertise + lease formation | Option retrieval + DNS behavior validation |
| Typical “first fix” | Fix RA flags to request DHCPv6 addresses | Fix DHCPv6 DNS option values and reachability |
| Default expectations in modern stacks (2026) | Still needs correct RA behavior | Relies on RA + DHCPv6 options working together |
| Verdict | Choose when you need centralized addressing control | Choose when RA is already reliable and you mainly need DNS/options |
7 Key IPv6 DHCPv6/RA Constants You Must Match (Practical 2026 Defaults)
| # | What to verify | Exact value | Why it matters | Priority |
|---|---|---|---|---|
| 1 | DHCPv6 client UDP port | 546 | Client sends solicit/request to the correct destination | ★★★★★ |
| 2 | DHCPv6 server UDP port | 547 | Server listens for advertise/reply traffic | ★★★★★ |
| 3 | ICMPv6 RA message type | 134 | RA drives DHCPv6 vs address autoconfig behavior | ★★★★★ |
| 4 | ICMPv6 RS message type | 133 | Clients/routers use RS to trigger RA faster | ★★★★☆ |
| 5 | DHCPv6 server/relay multicast (link-local) | ff02::1:2 | Defines where solicit/request messages are received | ★★★★☆ |
| 6 | Link-local IPv6 address prefix | fe80::/10 | Used for neighbor/RAs/DHCPv6 discovery on local links | ★★★★☆ |
| 7 | RA destination multicast | ff02::1 | Ensures clients receive RA to learn configuration | ★★★☆☆ |
Prepare prerequisites before touching settings
You reduce downtime by collecting router/server details and deciding what DHCPv6 should distribute before you change any RA or DHCPv6 flags. This matters because both RA and DHCPv6 are timing- and interface-sensitive—changing the wrong thing on the wrong interface is the fastest route to broken IPv6, even when the config “looks” correct.
– Gather basics: server/router model, OS/software version, the target interface name (e.g., `eth0` / `br0`), and the upstream prefix info.
– Decide what you will distribute via DHCPv6 (common options: DNS recursive resolver, domain search list, NTP, and domain name).
– Plan an easy test: pick one client device to validate first and keep logs accessible on the DHCPv6 server/router.
A practical rule: treat DHCPv6 configuration as two halves—(1) *plumbing* (correct interface, correct mode, correct prefix) and (2) *content* (DNS/NTP/etc.). If you can’t clearly identify those halves in your change plan, you’ll struggle to explain why connectivity works but DNS doesn’t.
The 3 numbers worth memorizing for IPv6 DHCP troubleshooting
When you’re debugging IPv6 DHCP in 2026, these constants help you sanity-check quickly:
– According to RFC 4861, RA timing uses a minimum interval of 3 seconds and a maximum of 10 seconds (MinRtrAdvInterval/MaxRtrAdvInterval).
– According to RFC 8415, DHCPv6 runs on UDP port 546 (client) and port 547 (server).
– According to RFC 4291, link-local addresses use the prefix fe80::/10, which is central to local-link discovery.
These values won’t fix misconfiguration by themselves, but they make it easier to interpret whether your client is seeing RA frequently enough to request DHCPv6 options (stateless) or to move into the correct lease behavior (stateful).
Configure the IPv6 DHCPv6 server on your router/server
You configure DHCPv6 successfully when you bind the DHCPv6 server to the LAN/client interface and set the correct mode-specific address/prefix behavior. Then you add DHCPv6 options like DNS so clients learn what they need without relying on local guesses.
– Enable DHCPv6 server/service on the correct interface (the one connected to clients, not the upstream WAN).
– Set the address/prefix scope carefully:
– For stateful operation, configure the IPv6 address pool/range that DHCPv6 should assign (or reference a delegated prefix with the intended allocation rules).
– For stateless operation, configure DHCPv6 to provide information-only (DNS/etc.) while RA manages addressing.
– Add DHCPv6 options clients need, especially:
– DNS servers (the recursive resolvers clients should use)
– Domain name / search list if you want consistent hostname resolution
– Optional: NTP, SIP server, or other site-specific options supported by your platform
From my day-to-day troubleshooting experience, the “looks correct but doesn’t work” symptom is usually one of these:
1) DHCPv6 is enabled on the WAN interface, so clients never receive responses.
2) Statefulness/prefix doesn’t align with what RA advertises, so clients refuse the lease or never request it.
3) DNS options are delivered but point to resolvers that aren’t reachable from the client subnet (firewall/routing issue).
Operational tip (2026-proofing): if your router supports multiple DHCPv6 profiles, create one dedicated profile for the LAN subnet and keep the upstream profile disabled. Less “helpful” automation reduces the chance of conflicting services.
Configure RA (Router Advertisements) to match your DHCPv6 plan
You should configure RA so the client’s interpretation of “where addresses come from” matches your DHCPv6 mode. If RA implies one behavior and DHCPv6 is configured for the other, clients will either never request what you’re offering or will treat your responses as irrelevant.
– Align RA flags with your intended DHCPv6 mode:
– If you want clients to use DHCPv6 for addresses, your RA/flags should reflect that stateful behavior.
– If you want RA to handle addressing, RA should provide addressing information and DHCPv6 should be limited to configuration info (like DNS).
– Ensure RA prefix settings match what DHCPv6 expects (especially if you’re using a delegated prefix).
– Validate RA parameters like prefix lifetime and advertisement intervals aren’t being overridden unexpectedly by another component (sometimes multiple services can advertise).
RA configuration is more than “turn it on.” In many router platforms, enabling DHCPv6 server also adds assumptions about RA flags, but those assumptions may be wrong for your chosen model. As a best practice, change RA first, then DHCPv6 options, and only then enable the address pool (if stateful).
A practical checklist for RA alignment
– Confirm the RA prefix length matches your addressing plan (most deployments use /64 for a LAN link).
– Ensure RA lifetime/validity isn’t effectively expiring faster than clients expect.
– Confirm RA is emitted on the LAN-facing interface where clients listen.
– Ensure no other process is also advertising prefixes on the same link (for example, if bridging/VPN services each enable RA).
Verify clients can actually obtain DHCPv6 settings
You validate IPv6 DHCP by checking two outcomes: (1) clients get the expected IPv6 address behavior and (2) clients use the DNS/options DHCPv6 provides. Even when DHCPv6 “works” in logs, DNS issues can still break real usage (email, web, updates).
– On the client, confirm it’s using IPv6 and has the expected address type:
– With stateful DHCPv6, clients should receive addresses from the DHCPv6 pool/range.
– With stateless DHCPv6, clients should already have an address from RA (and still learn DNS/other data via DHCPv6).
– On the DHCPv6 server/router, check logs for DHCPv6 solicit/advertise/lease messages (names vary by platform).
– Test the outcome that matters:
– Can the client reach an IPv6 host (basic connectivity)?
– Does the client resolve DNS using the DHCPv6-provided resolver (not just the router’s defaults)?
In my experience, the fastest way to separate address vs DNS problems is to run two tests back-to-back:
1) IPv6 connectivity test (ping an IPv6 literal or connect to an IPv6-only service).
2) DNS test (resolve a hostname that clearly requires DNS, then verify it’s using the intended resolver).
Also, if your router or client has IPv6 privacy addresses enabled, you may see address churn (stable behavior, different addresses). That’s fine—what’s not fine is missing or incorrect DNS.
What can go wrong (and how to spot it fast)
You can usually pinpoint the root cause within minutes by focusing on mode alignment, interface selection, and prefix/pool matching. When IPv6 DHCP breaks, it breaks in predictable ways—especially around RA vs DHCPv6 expectations.
– RA and DHCPv6 mode conflict: The most common failure is when RA tells clients to self-configure (or when flags don’t request DHCPv6) but you expect DHCPv6 to provide addresses.
– Wrong interface enabled: DHCPv6 server enabled on the WAN side instead of the LAN/client-facing side results in “nothing happens” from the client perspective.
– Address pool/prefix mismatch: In stateful mode, if the DHCPv6 pool doesn’t align with the prefix/expected subnet, clients may reject leases or never form usable connectivity.
– DNS looks broken even when DHCP works: Clients may successfully get DHCPv6 configuration but still fail name resolution if DNS addresses are incorrect, unreachable, or filtered.
– Client doesn’t request DHCPv6: Some environments (managed OS images, hardened configurations, or captive portal behaviors) may restrict DHCPv6 client activity. Check client network settings and IPv6 privacy/addressing preferences.
Quick pros/cons: choosing DHCPv6 complexity responsibly
| Consideration | Pros | Cons |
|---|---|---|
| Centralized DNS & options | Consistent resolver and search/domain settings for all clients | Still must ensure resolvers are reachable and firewalled correctly |
| Stateful address control | Predictable leases and easier auditing | More failure points if pools/prefixes don’t match RA |
| Troubleshooting visibility | DHCPv6 logs show solicit/advertise/lease progression | RA and DHCPv6 interaction can confuse “what’s actually failing” |
| Time sensitivity (2026) | Clients recover quickly when timers and lifetimes are reasonable | Overlapping services can advertise competing lifetimes |
Verdict: use DHCPv6 when you want central control—skip it if your goal is only basic IPv6
IPv6 DHCP (DHCPv6 + RA alignment) is worth it when you need centralized, repeatable distribution of DNS and other network settings, or when you explicitly want stateful address assignment. The tradeoff is complexity: RA behavior, DHCPv6 options, and pool/prefix configuration must agree, and troubleshooting can take longer than IPv6 stateless-only setups.
Skip DHCPv6 (or keep it minimal) if:
– You only need clients to autoconfigure addresses and you can live with static DNS configuration elsewhere.
– You’re not ready to troubleshoot RA/DHCPv6 interaction when connectivity fails.
If your environment is small and relatively stable, stateless DHCPv6 (DNS/options only) is often the lowest-risk way to get centralized name resolution without turning every client’s addressing into a DHCPv6 lease workflow. As of late 2026, that approach remains common in home lab and small office deployments because it reduces moving parts while still solving one of the most visible pain points: DNS consistency.
Quick checklist (scan before you commit changes)
| Check | What “good” looks like |
|---|---|
| DHCPv6 mode chosen | Stateful (addresses) or stateless (info-only) is intentional |
| Correct interface | Enabled on LAN/client-facing interface, not upstream WAN |
| RA matches intent | RA flags/timers align with stateful vs stateless DHCPv6 behavior |
| Prefix/address alignment | DHCPv6 pool/prefix matches RA prefix expectations |
| DHCPv6 options set | DNS (primary/secondary), domain/search, and any required extras |
| Client validation | Client gets usable IPv6 address + correct DNS behavior |
Frequently Asked Questions
How do I configure IPv6 DHCP on a home or small business router?
Start by enabling the DHCPv6 server feature in your router’s IPv6 settings and specify the LAN interface that will hand out addresses. Choose the correct mode (usually “stateful” if you want clients to receive full addresses and DNS, or “stateless” if only configuration flags are needed). Then set the IPv6 prefix delegation (if your ISP provides it) and configure DNS servers (router as recursive resolver or specified upstream DNS). Finally, save and reboot or apply changes, then test on a client by releasing and renewing the IPv6 DHCP lease.
What is the difference between stateful and stateless IPv6 DHCP, and which should I use?
Stateful DHCPv6 assigns clients complete IPv6 addresses along with additional options like DNS and domain name, similar to IPv4 DHCP. Stateless DHCPv6 typically provides only configuration information (such as DNS server addresses) while clients still form their own addresses using SLAAC from router advertisements. If you need consistent address assignment and centralized control, stateful DHCPv6 is often the better choice; if you prefer SLAAC-based addressing with minimal management, stateless may fit better. Many networks use a combination with router advertisements for addressing and DHCPv6 for DNS and other options.
How do I set up IPv6 DHCP options like DNS, domain name, and NTP servers?
In your DHCPv6 configuration, look for “DHCP options” or “IPv6 DHCP settings” and add DNS (Recursive name servers) addresses and any domain name options if your devices rely on them. For NTP, enable DHCPv6 NTP options and provide the NTP server addresses or the appropriate option fields supported by your router or DHCPv6 server software. Verify that your clients support the options you enable—some operating systems may use SLAAC-provided DNS unless DHCPv6 options override it. After applying settings, test with a client using commands such as checking IPv6 DNS server assignment on Windows/macOS/Linux.
Which IPv6 DHCP mode and settings should I use when my ISP provides prefix delegation (PD)?
When using IPv6 prefix delegation, you typically delegate an upstream prefix to your router, then use that prefix to define the downstream DHCPv6 address pool and (optionally) router advertisement prefixes. For stateful DHCPv6, configure a DHCPv6 address pool that falls within the delegated prefix and ensure the prefix length and range match your LAN subnet design. For stateless operation, you’ll still need router advertisements to advertise the prefix, while DHCPv6 can supply DNS and other parameters. Confirm that your PD renew behavior updates the DHCPv6 pool correctly so clients keep connectivity after prefix changes.
What are common IPv6 DHCPv6 troubleshooting steps when clients don’t get addresses or DNS?
First, verify the DHCPv6 server is enabled on the correct interface and that the IPv6 prefix advertised or pooled by the server is valid and reachable. On clients, confirm they are requesting DHCPv6 (and not only using SLAAC) and check for duplicate or conflicting settings like multiple IPv6 routers advertising. If DNS isn’t working, validate that DHCPv6 options for DNS are actually configured and that clients are configured to use them; you may need to adjust whether DNS is provided by SLAAC RDNSS or by DHCPv6. Finally, check router/firewall rules for UDP ports used by DHCPv6 (typically server-side port 546/547 and client-side port 547) and review logs for DHCPv6 solicit/advertise events.
📅 Last Updated: November 17, 2026 | Topic: How to Configure IPv6 DHCP | Content verified for accuracy and freshness.
References
- https://en.wikipedia.org/wiki/Dynamic_Host_Configuration_Protocol_for_IPv6
- RFC 8415: Dynamic Host Configuration Protocol for IPv6 (DHCPv6) | RFC Editor
- RFC 3633: IPv6 Prefix Options for Dynamic Host Configuration Protocol (DHCP) version 6 | RFC Editor
- RFC 3646: DNS Configuration options for Dynamic Host Configuration Protocol for IPv6 (DHCPv6) | R…
- RFC 3315: Dynamic Host Configuration Protocol for IPv6 (DHCPv6) | RFC Editor
- RFC 3319: Dynamic Host Configuration Protocol (DHCPv6) Options for Session Initiation Protocol (S…
- Dynamic Host Configuration Protocol for IPv6 (DHCPv6)
- https://scholar.google.com/scholar?q=IPv6+DHCPv6+configuration+guide Google Scholar
- https://scholar.google.com/scholar?q=DHCPv6+prefix+delegation+IA_PD+configuration Google Scholar
- https://scholar.google.com/scholar?q=RFC8415+DHCPv6+implementation+configuration Google Scholar