How Do You Make DNS Keep Working When the Internet Is Down?

Keep local DNS local. Your router, Pi-hole, AdGuard Home, Technitium, or internal resolver should answer local names without forwarding to the internet, and clients should have at least two reachable resolvers that behave consistently.

Design principle: Make the network boring on purpose: clear ownership, few trust zones, documented DNS, and access paths that fail closed.

The Short Version

• Keep local DNS local. Your router, Pi-hole, AdGuard Home, Technitium, or internal resolver should answer local names without forwarding to the internet, and clients should have at least two reachable resolvers that behave consistently.
• Use the decision matrix below, then prove the result with the validation checklist before making it the default.

Why This Matters Now

The useful answer starts with the operating model. Who depends on this service, what breaks when it is unavailable, and how quickly does it need to be restored? Those questions matter more than the product name.

Many homelabs fail strangely during WAN outages because local names depend on public DNS, external proxy records, or cloud identity.

DNS should be treated as infrastructure with redundancy, monitoring, and a failure mode.

The goal is not just ad blocking; it is predictable name resolution for local services.

The rest of this guide turns that context into a baseline design, implementation order, validation checks, and buying notes. That is the TechGeeks bias: a setup is not good because it worked once. It is good when it can be explained, tested, and recovered.

Recommended Baseline

Start with ownership. One device should own routing and firewall policy, one plan should define DNS, and each VLAN or SSID should exist because a trust boundary changed. If two systems are both trying to be DHCP, DNS, VPN gateway, or reverse proxy, the network will eventually become harder to debug than it needs to be.

The baseline is simple: documented subnets, named infrastructure addresses, router configuration backups, local DNS that survives WAN trouble, and remote access that starts private unless a service truly needs public users.

Decision Matrix

Choice	Best Fit	Watch Point
Router DNS only	Small simple networks.	Weak logging and policy.
Pi-hole/AdGuard	Filtering plus local records.	Needs HA plan if critical.
Technitium/Unbound	More DNS control.	More complexity.
Active Directory DNS	Windows domain environments.	Heavy for normal homes.

Decision Worksheet

Before copying the recommendation, fill out this worksheet for your own home or lab. The right answer can change when the same tool is used for family photos, router access, media playback, cameras, or a disposable test stack.

Worksheet Item	What To Write Down	Why It Matters
Primary question	How do I make DNS keep working when the internet is down?	This keeps the article tied to the reader's real decision instead of drifting into a generic product comparison.
Affected systems	The devices and services that lose internet, DNS, Wi-Fi, remote access, or admin reachability if this fails.	Readers should know who and what they are protecting before they choose hardware, software, or a cloud service.
Failure model	WAN outage, bad DNS, blocked discovery, stale firewall rules, expired certificates, and lost admin access.	Different failures need different controls. This row prevents RAID, sync, VPN, or MFA from being treated as magic.
Proof test	Test from a wired client, Wi-Fi client, phone on cellular, and any VLAN or tunnel that depends on the change.	A recommendation is not proven until it survives a small, repeatable test using realistic data, clients, or accounts.
Rollback path	Export config first and identify the old port, SSID, DNS server, or tunnel setting that restores service.	A reversible change is less stressful, easier to explain, and less likely to turn a weekend project into an outage.
Measurement to capture	Latency and throughput from the rooms or VLANs that matter, not just beside the router.	Numbers, logs, screenshots, or restore notes give the reader confidence that the decision was based on evidence.

Run A WAN-Disconnect Drill

Unplug the WAN and watch what still works. Local names under home.arpa should resolve if your local resolver is healthy. Cached public names may work briefly. New public lookups should fail cleanly. DHCP leases should renew. NTP may drift if there is no local source. Apps that hard-depend on cloud identity may fail even when local DNS is fine.

Keep DNS HA simple. Two resolvers help only if clients know both, both have the same local records, and the router does not hand out a dead upstream path. Document the emergency bypass before an outage.

Real-World Example

Consider a home where the router, NAS, Home Assistant, media server, and family laptops all depend on one flat network. The better design is a small number of understandable trust zones, a DNS path that still works during WAN trouble, and remote access that starts private by default. Success is not a prettier dashboard; success is being able to explain which device can reach which service and why.

Draw the path for one real workflow from start to finish. For example: phone on Wi-Fi, DNS resolver, firewall rule, reverse proxy or tunnel, application container, database, and storage mount. Then repeat it from a phone on cellular if remote access is part of the design. That path exposes the hidden dependencies that a feature comparison misses.

The practical lesson is that most network problems are ownership problems. One system should own routing, one plan should define DNS, and each trust boundary should have written rules. If the reader cannot explain where DHCP, DNS, firewall policy, and remote identity live, the next outage will feel random even when the tools are working as designed.

Rollout And Recovery Plan

Treat network changes like small production changes. Export the router or firewall configuration, write down the current DNS and DHCP settings, and keep one known-good admin path available while you test. If the change involves VLANs, tunnels, reverse proxies, or DNS policy, move one noncritical client first instead of changing the whole house at once.

The rollback plan should be boring: which config backup to restore, which cable or port returns a device to the old network, which DNS server bypasses the new resolver, and which hostname or tunnel can be disabled quickly. If you cannot describe rollback in one paragraph, the change is probably too broad for one maintenance window.

Implementation Details

Implement this in a maintenance window, even if the word maintenance feels too formal for a home lab. The point is to avoid changing several hidden dependencies while someone else expects the internet, photos, media, smart home, or passwords to keep working.

1. Choose a local domain that does not collide with public domains you do not own.
2. Put important local records in the resolver clients actually use.
3. Use DHCP options to hand out resolver IPs consistently.
4. Avoid giving clients an ISP resolver as secondary if it breaks local names.
5. Document how to bypass the filtering resolver during an outage.

Record these details while you build, not after the memory has already gone fuzzy:

• Latency and throughput from the rooms or VLANs that matter, not just beside the router.
• DNS behavior when the WAN is unplugged, VPN is connected, and browser secure DNS is enabled.
• Firewall logs for denied traffic between guest, IoT, management, and trusted networks.
• Open ports and externally reachable hostnames after the change.

Evidence To Collect

The article should leave the reader with something they can verify. Collecting evidence sounds formal, but it can be as small as a restored folder, a router config export, a playback dashboard capture, or a clean-browser login test.

• Current router, firewall, switch, access point, and DNS configuration exports before the change.
• Client evidence from the actual device: IP address, gateway, DNS servers, VLAN or SSID, and browser secure-DNS state.
• A test from outside the house, preferably cellular, when remote access or public exposure is part of the design.
• Firewall, tunnel, proxy, and DNS logs that show both allowed traffic and expected denies.
• A list of open ports, public hostnames, certificate expiry dates, and stale VPN or tailnet devices.

Failure Signals

• Local names stop working when the internet is down.
• Clients randomly use different DNS servers or bypass policy with browser secure DNS.
• Admin pages are reachable from guest, IoT, or public networks.
• No one can describe which device owns routing, DHCP, DNS, and remote access.

Adopt, Pilot, Defer, Avoid

• Adopt: Adopt the network change when ownership, DNS, firewall policy, remote access, and rollback are documented.
• Pilot: Pilot with one client, one VLAN, one hostname, or one tunnel before moving the whole house.
• Defer: Wait when the current setup is stable, backed up, monitored, and the proposed change is mostly curiosity.
• Avoid: Avoid exposing admin interfaces or broad internal networks just because a tunnel or reverse proxy makes it convenient.

Validation Checklist

• Local service names resolve with WAN disconnected.
• Router, NAS, Proxmox, Home Assistant, and reverse proxy names resolve from trusted devices.
• Both DNS servers return the same local answers.
• Clients receive expected DNS servers from DHCP.
• Monitoring alerts if one resolver goes down.

Common Mistakes

• Using ISP DNS as secondary and wondering why local names randomly fail.
• Depending on public DNS records for private services during WAN failure.
• Running DNS on the same host that depends on DNS to boot.
• Changing browser DoH settings without documenting them.
• Letting VPN clients use different DNS rules than home clients.

Troubleshooting

Symptom	Likely Cause	First Check
Clients behave differently	DHCP, browser secure DNS, VPN DNS, IPv6, or manual settings are bypassing policy.	Check the resolver and gateway from the actual client, not only from the router UI.
Remote access breaks	Identity, DNS, tunnel routing, firewall policy, or certificate renewal changed.	Test from a mobile hotspot and review logs at the tunnel, proxy, and app layers.
Segmentation breaks apps	Discovery or controller traffic was blocked along with broad LAN access.	Add narrow mDNS, controller, DNS, NTP, or app-port exceptions and document them.

Maintenance Cadence

The best design is the one that still makes sense three months later. Put these checks on a calendar so the setup does not depend on memory.

• Monthly: Review firmware, open ports, DNS failures, VPN users, certificate expiry, and noisy firewall blocks.
• Quarterly: Run a WAN-disconnect or remote-access test and confirm local names, admin access, and rollback notes still work.
• Yearly: Audit network segmentation, retire stale devices, and confirm router or firewall backups restore to current hardware.

Network maintenance should include a failure drill. Unplug WAN, test remote access from cellular, confirm local DNS, and verify that the config export is stored somewhere other than the router or firewall.

When To Spend Money

Product links make sense only after the reader knows what problem the purchase solves. Use this table to keep buying advice tied to evidence, not anxiety or a tempting sale price.

Stage	Signal	Practical Buying Guidance
Do not buy yet	Coverage, DNS behavior, firewall policy, and client path have not been measured.	Map the network, export configs, test clients, and identify the bottleneck first.
Small useful spend	The design is sound but lacks one reliable link, management path, or recovery aid.	Managed switch, spare patch cables, labels, UPS for network gear, or a travel router for remote access testing.
Larger upgrade	Measured throughput, segmentation, VPN, Wi-Fi coverage, or routing limits block a real workflow.	Firewall appliance, access points with wired backhaul, 2.5GbE/10GbE switch, or a supported router platform.

Useful Gear And Buyer Notes

The product links below are intentionally search links, starting with Raspberry Pi 5 kit, because model numbers, bundles, and prices change quickly. Use them to compare categories, then verify exact specifications against the article's decision points before buying. For infrastructure gear, prioritize firmware support, replaceability, warranty, idle power, and recovery behavior over headline specs.

Affiliate disclosure: As an Amazon Associate, TechGeeks may earn from qualifying purchases. The product links below are buying references, not a requirement to buy a specific brand or seller. Verify compatibility, seller quality, warranty, and current specs before ordering.

• Amazon search: Raspberry Pi 5 kit
• Amazon search: Intel N100 mini PC
• Amazon search: 2.5GbE mini PC firewall
• Amazon search: UPS small network
• Amazon search: managed switch

Related TechGeeks resources

• Homelab VLAN Design: Simple Network Segmentation That Works
• IoT Isolation for Homelabs: VLANs, Firewall Rules, and mDNS
• WireGuard Home VPN: Secure Remote Access for Your Homelab
• Homelab DNS Guide: Local Names, Ad Blocking, and Reliability

What This Does Not Protect or Validate

This guide does not guarantee that vendor pricing, product bundles, firmware behavior, subscription terms, or cloud policies will stay the same. Verify current documentation before final buying or migration decisions.

It also does not replace a full security, backup, or disaster-recovery program. The goal is to give you a practical design, the tests that prove it, and the boundaries that keep the recommendation honest.

Segmentation, VPNs, tunnels, DNS filtering, and reverse proxies reduce risk only when firewall rules, logs, updates, and account recovery are maintained.

Practical FAQ

How do I make DNS keep working when the internet is down?

Keep local DNS local. Your router, Pi-hole, AdGuard Home, Technitium, or internal resolver should answer local names without forwarding to the internet, and clients should have at least two reachable resolvers that behave consistently. The important next step is to validate the recommendation with one small test before treating it as the default.

References

• https://pi-hole.net/
• https://github.com/AdguardTeam/AdGuardHome
• https://technitium.com/dns/
• https://nlnetlabs.nl/projects/unbound/about/
• https://www.rfc-editor.org/rfc/rfc6762

Final Thought

The right answer is the one you can operate, document, test, and recover without guessing.