Nerd Family – WiFi IPv6-Only Network Connection Fail
Warning — Should be addressed soon
Symptoms
- Miner UI / OLED / TFT shows `WiFi connected` with strong RSSI (-40 to -65 dBm) but hashrate stays at 0
- Pool subscribe / authorize never completes; dashboard pool field stuck on `connecting...`
- Serial console at 115200 baud shows DNS errors: `getaddrinfo failed`, `lwip_getaddrinfo: 202` (`EAI_FAIL`), `dns_gethostbyname: 202`, `socket connect: errno 113` (`EHOSTUNREACH`)
- Pool URL field shows hostname (`solo.ckpool.org`, `public-pool.io`, etc.) but never resolves to an IP
- Router admin page shows the miner with only IPv6 (link-local `fe80::...` or global `2001:...`) and no IPv4 lease, or only an APIPA `169.254.x.y` address
- Active carrier is T-Mobile US Home Internet 5G, EE 5G Smart Hub IPv6, Vodafone DE IPv6 mobile APN, Bouygues Bbox IPv6, Bell or Rogers IPv6 trial, or a similar 5G FWA / mobile network
- Laptops and phones on the same SSID can browse the web normally (because they do `464XLAT` / NAT64 / DNS64 in the OS — the miner cannot)
- Miner works at a friend's house / office and not at home, or vice versa (geographic / network dependency)
- Symptom started after switching ISPs, moving to a 5G home gateway, or moving into a building with a mandated carrier router
- Symptom timed with mobile-tower IPv6 rollout in your region (EU IPv6-only APN rollout 2024-2026; North American 5G FWA gradual IPv4 deprecation)
- On Bitaxe Web Flasher console: `mining.subscribe` never sends, even though Wi-Fi association banner is green
- Carrier gateway admin page shows no IPv4 LAN configured at all, or only a `100.64.0.0/10` CGNAT-style IPv4 with an aggressive deprecation timer
Step-by-Step Fix
Confirm the network is IPv6-only. From a laptop or phone on the same Wi-Fi, open `test-ipv6.com` or `ip.tyil.nl/ipv6`. If the page reports `no IPv4 connectivity / IPv6 only`, the LAN is IPv6-only and the miner is hitting the LWIP gap. If both protocols are reported reachable, the failure is something else — re-diagnose against the related-error list.
Substitute a phone hotspot for 10 minutes. Tether on a normal IPv4 mobile plan (NOT a T-Mobile Home Internet SIM, NOT an EE 5G Home SIM). Connect the miner. If it hashes within 60 seconds, the home network is IPv6-only and the miner is healthy — focus on the network, not the device.
Restart the carrier gateway. Some IPv6-only deployments occasionally hand out a temporary IPv4 lease during prefix-delegation glitches. A reboot may give a 24-hour mining window while you set up a permanent fix. This is a band-aid, not a solution.
Move the miner to a guest SSID on a router upstream of the IPv6-only carrier link. If you have your own router doing dual-stack NAT44 before the carrier gateway became authoritative, plug the miner there. Most homes won't have this topology, but some do (carrier modem in bridge mode + your own router).
Change the pool URL to D-Central's stratum endpoint or a community pool with documented dual-stack DNS (publishes both A and AAAA). Some pools resolve cleanly even on IPv6-only LANs; most don't. Lowest-effort experiment before touching DNS settings.
Override LAN-side DNS at the carrier gateway to a dual-stack public resolver: Google Public DNS (`8.8.8.8`, `8.8.4.4`, `2001:4860:4860::8888`, `2001:4860:4860::8844`) or Cloudflare (`1.1.1.1`, `1.0.0.1`, `2606:4700:4700::1111`, `2606:4700:4700::1001`). The carrier's own DNS is often DHCPv6-only, which ESP32 LWIP doesn't reliably consume. Save, reboot the miner, watch the dashboard.
Use Google Public DNS64 / Cloudflare DNS64 if Step 6 doesn't fix it: `2001:4860:4860::6464`, `2001:4860:4860::64`, `2606:4700:4700::6400`, `2606:4700:4700::64`. These return synthetic AAAA records for IPv4-only hosts, letting the miner reach pool hostnames over NAT64. Works only if the carrier's NAT64 itself is stable.
Reserve a static DHCP lease for the miner where the gateway lets you. Even IPv6-only deployments often allow a parallel `192.168.x.0/24` IPv4 LAN, just disabled by default on 5G FWA gateways. Some T-Mobile Arcadyan / Sagecom and EE Smart Hub units expose this behind an `advanced` toggle; enabling it gives the miner a dual-stack home and the LWIP gap disappears.
Check whether your firmware was built with IPv6 enabled. Some AxeOS / NerdMiner / NerdNOS builds disable IPv6 entirely in `sdkconfig` (`LWIP_IPV6=n`). On those builds, the miner refuses to associate with an IPv6-only SSID at all — the symptom is `Wi-Fi never connects`, not `Wi-Fi connects but no pool`. Cross-check by flashing a build with documented IPv6 support and re-testing.
Switch carrier from IPv6-only mobile / FWA to fixed-line broadband for the miner. EU territory served by IPv6-only mobile APNs and US 5G FWA carriers are great for general internet but hostile to embedded mining firmware. Wired DSL / cable / FTTH plans almost universally still provide IPv4. One-time, permanent fix; operationally cheaper than fighting NAT64.
Drop an IPv4-only Wi-Fi bubble between the IPv6-only carrier and the miner. GL.iNet GL-MT300N-V2 (~$25 CAD), GL-AR300M (~$45 CAD), or any old OpenWRT-capable router. Plug into carrier gateway Ethernet, set WAN to DHCP, set LAN to `192.168.50.0/24` IPv4 NAT44, run a dedicated SSID like `mining-net`, connect every Nerd / Bitaxe to it. Carrier absorbs the IPv6 complexity; miners see a normal dual-stack-compatible IPv4 LAN. D-Central recommended fix.
Configure NAT64 / DNS64 on your own router (OpenWRT). Install `464xlat`, `dns64`, or `nat64` packages, point LAN DNS at the local `dns64` daemon, let it synthesize AAAA on the fly. Works as a single-router solution but operationally more complex than Step 11. Recommend only if you're already running OpenWRT for other reasons.
Build a firmware with explicit dual-stack support (`LWIP_IPV6=y`, `LWIP_IPV6_DHCP6=y`, `LWIP_IPV6_RDNSS=y`, `LWIP_DNS=y` in `sdkconfig`) and flash it. Requires PlatformIO / ESP-IDF and the relevant repo (`bitaxeorg/ESP-Miner`, `BitMaker-hub/NerdMiner_v2`, `NerdMiner-org/NerdNOS`, `Shufflepuck/NerdQAxe-plus`). Closes the firmware-side gap; NAT64 reachability still depends on the carrier.
Run a Stratum proxy on a $5/month VPS (Hetzner, Vultr, DigitalOcean) with a clean dual-stack uplink, plus a small `stratum-proxy` daemon on the VPS public IP. Point miner at VPS hostname. Proxy translates between miner's local connection and pool's IPv4 endpoint, side-stepping carrier NAT64. Same approach with Cloudflare Tunnel / WireGuard / Tailscale exit-node. Use when carrier blocks pool ports outright.
Test pool reachability from inside the IPv6-only LAN with `dig` / `nslookup` / `nc`. From any phone or laptop on the same SSID: `dig solo.ckpool.org A`, `dig solo.ckpool.org AAAA`, `nc -vz solo.ckpool.org 3333`. Output reveals exactly which layer (DNS, AAAA synthesis, TCP reach) is broken. Bring that output to D-Central support — diagnostic time becomes near-zero.
Stop DIY when the IPv4-only travel router is in place, dual-stack DNS confirmed on a laptop, pool resolves and pool port reachable from a laptop on the same LAN — and the miner still can't reach a pool. The failure is no longer the LAN; it's a device-side fault (firmware corruption, ESP32-S3 fault, hashboard fault). Book D-Central's troubleshooting bench.
D-Central bench process. We replicate the IPv6-only-network setup (a labelled T-Mobile Home Internet test gateway and an EE 5G test SIM live in our network closet), capture the miner's serial logs end-to-end through the failure, reproduce the resolver / NAT64 / port chain, and either confirm a network fix you can replicate at home or identify a genuine device fault. Most cases come back as `router config` with a short report.
Ship safely. Anti-static bag, double-box with at least 5 cm foam every side. Include a note with: model + revision, firmware build hash, exact ISP / carrier / gateway model, screenshots of the gateway admin page showing the miner's lease state, serial console capture, and the laptop-side `dig` / `nc` test results from Step 15. Complete network description shaves an hour of bench time off your invoice.
When to Seek Professional Repair
If the steps above do not resolve the issue, or if you are not comfortable performing these repairs yourself, professional service is recommended. Attempting advanced repairs without proper equipment can cause further damage.
Related Error Codes
Still Having Issues?
Our team of Bitcoin Mining Hackers has been repairing ASIC miners since 2016. We have seen it all and fixed it all. Get a professional diagnosis.
