fix(server): prevent zfs-init-tank from wiping pool on udev race

zpool create -f was called if `zpool import -d /dev/disk/by-id` failed,
which could happen due to a race with systemd-udev-settle. The disk
would then be visible by the time zpool create ran, silently destroying
all data on an otherwise-intact pool.

Fix: locate the data disk first, retry the import directly against it
as a fallback, then check zdb -l for existing ZFS label metadata before
concluding the disk is blank. Remove -f so zpool create refuses rather
than overwrites if a pool is present.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01ULXzafSDwGhmFGnn3LtDSQ

.sops.yaml

@@ -77,6 +77,40 @@ creation_rules:
           - *admin
           - *lxc-tor-relay
 
+  - path_regex: secrets/tailscale-router\.yaml$
+    key_groups:
+      - age:
+          - *admin
+          - *linode-tailscale-router
+          - *lxc-tailscale-router
+          - *proxmox-tailscale-router
+
+  # HA file server per-node secrets (beszel-token).
+  # proxmox-ha-server-1 / proxmox-ha-server-2 keys are added automatically
+  # by scripts/secrets/sync-host-keys.sh once the hosts are provisioned;
+  # until then only the admin key can decrypt these files.
+  - path_regex: secrets/ha-server-1\.yaml$
+    key_groups:
+      - age:
+          - *admin
+          # proxmox-ha-server-1 added by sync-host-keys.sh
+
+  - path_regex: secrets/ha-server-2\.yaml$
+    key_groups:
+      - age:
+          - *admin
+          # proxmox-ha-server-2 added by sync-host-keys.sh
+
+  # Shared HA cluster corosync authkey (binary sops file).
+  # Encrypted for both HA nodes so either can decrypt on boot.
+  # Both host keys added by sync-host-keys.sh; admin key allows initial creation.
+  - path_regex: secrets/ha-corosync-authkey$
+    key_groups:
+      - age:
+          - *admin
+          # proxmox-ha-server-1 added by sync-host-keys.sh
+          # proxmox-ha-server-2 added by sync-host-keys.sh
+
   # gui-host-specific secrets (currently: wifi-password, see
   # modules/networking/wifi.nix). Only *lxc-gui has a registered key today
   # -- proxmox-gui/linode-gui/baremetal-gui haven't been provisioned via

docs/ip-addressing.md

@@ -0,0 +1,128 @@
+# IP Addressing Scheme
+
+## Subnets
+
+| Subnet | CIDR | Purpose | Routed? |
+|---|---|---|---|
+| LAN | `192.168.2.0/24` | General LAN — clients and infrastructure | Yes (gateway .254) |
+| Storage | `192.168.4.0/29` | HA file server DRBD replication | No — internal `vmbr1` only, no uplink |
+
+The storage subnet never leaves pve1. `vmbr1` is a Proxmox Linux bridge with no physical port
+attached; traffic between the two HA file server VMs stays in-kernel.
+
+The host octet is consistent across subnets for any host that has multiple interfaces — e.g.
+ha-node1 is always `.228` (LAN: `192.168.2.228`, storage: `192.168.4.228`).
+
+---
+
+## LAN — 192.168.2.0/24
+
+### Address map
+
+| Range | Purpose |
+|---|---|
+| .1–.9 | Reserved, never assign |
+| .10–.59 | Client DHCP pool (router-assigned) |
+| .60–.219 | Unallocated buffer |
+| .220–.229 | Virtual nodes (VMs / LXC containers) |
+| .230–.239 | Expansion buffer (reserved, unallocated) |
+| .240–.249 | Physical nodes (bare-metal hosts) |
+| .250–.253 | Network services |
+| .254 | Router / gateway |
+
+### Network services (.250–.253)
+
+| IP | Hostname | Role |
+|---|---|---|
+| `192.168.2.254` | router | Gateway (TP-Link) |
+| `192.168.2.253` | domain-controller | FreeIPA — authoritative DNS for `sweet.home`, Kerberos, LDAP |
+| `192.168.2.250`–`.252` | — | Reserved for future network services |
+
+### Physical nodes (.240–.249)
+
+| IP | Hostname | Role |
+|---|---|---|
+| `192.168.2.245` | pve1 | Proxmox VE hypervisor |
+| `192.168.2.244` | pbs | Proxmox Backup Server |
+| `192.168.2.243` | nixos | Bare-metal workstation (`baremetal-gui`) |
+| `192.168.2.246`–`.249` | — | Reserved — second Proxmox node and associated services |
+| `192.168.2.240`–`.242` | — | Reserved |
+
+pve1 sits mid-range deliberately so a second Proxmox node can slot in on either side.
+
+### Virtual nodes (.220–.229)
+
+All VMs and LXC containers run on pve1.
+
+| IP | Hostname | Role | Status |
+|---|---|---|---|
+| `192.168.2.229` | ha-vip | HA file server iSCSI floating VIP (Pacemaker) | Future |
+| `192.168.2.228` | ha-node1 | HA file server node 1 (DRBD + XFS + iSCSI) | Future |
+| `192.168.2.227` | ha-node2 | HA file server node 2 (DRBD + XFS + iSCSI) | Future |
+| `192.168.2.226` | server | Current NFS/ZFS file server — retires when HA is live | Retiring |
+| `192.168.2.225` | docker | Docker / Traefik stack | Active |
+| `192.168.2.224` | nix-cache | Nix binary cache + remote builder | Active |
+| `192.168.2.223` | pxe-boot | PXE / TFTP / HTTP netboot server | Active |
+| `192.168.2.222` | tailscale-router | Tailscale exit node / router | Active |
+| `192.168.2.221` | tor-relay | Tor relay | Active |
+| `192.168.2.220` | pdm | Proxmox Deploy Manager | Active |
+
+### Client DHCP pool (.10–.59)
+
+Assigned by the router. DNS option points to `192.168.2.253` (domain-controller).
+
+Devices in this range: phones, laptops, IoT, Canon printer, any non-infrastructure host.
+No static reservations for infrastructure hosts — all infra uses static IP configuration
+on the guest itself (not DHCP reservations), so IPs survive VM recreation regardless of
+MAC address churn.
+
+---
+
+## Storage network — 192.168.4.0/29
+
+Internal to pve1 only. Proxmox bridge `vmbr1`, no physical NIC attached.
+
+| IP | Hostname | Interface role |
+|---|---|---|
+| `192.168.4.228` | ha-node1 | DRBD replication NIC |
+| `192.168.4.227` | ha-node2 | DRBD replication NIC |
+| — | no gateway | Isolated — not routed to LAN or internet |
+
+---
+
+## Migration reference
+
+Current → target IP for every host being renumbered.
+
+| Host | Current IP | New IP | Config location |
+|---|---|---|---|
+| router | `192.168.2.254` | `192.168.2.254` | unchanged |
+| domain-controller | `192.168.2.138` | `192.168.2.253` | `/etc/sysconfig/network-scripts/ifcfg-eth0` on guest |
+| pve1 | `192.168.2.250` | `192.168.2.245` | `/etc/network/interfaces` on Proxmox host |
+| pbs | `192.168.2.108` | `192.168.2.244` | static config on PBS host |
+| nixos workstation | `192.168.2.119` | `192.168.2.243` | `networking.interfaces` / NetworkManager on guest |
+| ha-node1 | — | `192.168.2.228` | future |
+| ha-node2 | — | `192.168.2.227` | future |
+| ha-vip | — | `192.168.2.229` | future (Pacemaker resource) |
+| server | `192.168.2.252` | `192.168.2.226` | static config on guest |
+| docker | `192.168.2.249` | `192.168.2.225` | static config on guest |
+| nix-cache | `192.168.2.120` | `192.168.2.224` | static config on guest |
+| pxe-boot | `192.168.2.247` | `192.168.2.223` | static config on guest; update `vars.pxeServerIp` in `variables.nix` ✓ |
+| tailscale-router | `192.168.2.121` | `192.168.2.222` | static config on guest |
+| tor-relay | `192.168.2.107` | `192.168.2.221` | static config on guest |
+| pdm | `192.168.2.248` | `192.168.2.220` | static config on guest |
+
+### Cutover notes
+
+- **Do domain-controller first** — it becomes the DNS server; everything else depends on it
+  having its new IP and FreeIPA DNS configured before Pi-hole is retired.
+- **pve1 last among physical hosts** — changing the Proxmox management IP drops the web UI
+  briefly; all guests keep running.
+- **Update Pi-hole custom.list / FreeIPA DNS A records** to new IPs before flipping any host,
+  so name resolution stays valid throughout the migration.
+- **variables.nix already updated** for `pxeServerIp` (.247→.223), `pbsIp` (.108→.244), and
+  new `domainControllerIp` (.253). Rebuild affected hosts after renumbering.
+- **Router DHCP**: once domain-controller is at .253 and FreeIPA DNS is serving `sweet.home`,
+  switch router DHCP on with pool .10–.59 and DNS option pointing to .253; retire Pi-hole CT.
+- **Pi-hole's iPXE dnsmasq config** (`99-ipxe-chainload.conf`) moves to the pxe-boot CT as a
+  dnsmasq proxy-mode config before Pi-hole is decommissioned.

docs/network-cutover.md

@@ -0,0 +1,366 @@
+# Network Cutover Plan
+
+Moves the LAN from the current flat/Pi-hole-managed state to the new IP scheme
+defined in `docs/ip-addressing.md`. Works in five independent stages — each
+stage is safe to pause after and resume later. Rollback steps are given at
+every point where something can break.
+
+**Before starting anything:** confirm you have
+- SSH access to `192.168.2.138` (domain-controller, current IP)
+- SSH access to `192.168.2.250` (pve1)
+- Browser access to Pi-hole admin at `http://192.168.2.253`
+- Browser access to router admin at `http://192.168.2.254`
+- The FreeIPA `admin` password to hand
+
+---
+
+## Stage 1 — Prepare FreeIPA DNS (zero downtime)
+
+Everything here is additive. Pi-hole keeps running. Nothing breaks if you stop
+mid-stage.
+
+### 1a. Add NextDNS forwarders
+
+```bash
+ssh wayne@192.168.2.138
+kinit admin          # enter FreeIPA admin password when prompted
+ipa dnsconfig-mod \
+  --forwarder=45.90.28.142 \
+  --forwarder=45.90.30.142 \
+  --forward-policy=only
+```
+
+**Verify external resolution works through FreeIPA before continuing:**
+```bash
+dig @127.0.0.1 google.com +short   # must return an IP, not SERVFAIL
+```
+
+### 1b. Add A records for every host at their CURRENT IPs
+
+These represent the live state now. You'll update each record to the new IP
+when you renumber that host in Stage 5.
+
+```bash
+ipa dnsrecord-add sweet.home pve1             --a-rec 192.168.2.250
+ipa dnsrecord-add sweet.home pbs              --a-rec 192.168.2.108
+ipa dnsrecord-add sweet.home nixos            --a-rec 192.168.2.119
+ipa dnsrecord-add sweet.home server           --a-rec 192.168.2.252
+ipa dnsrecord-add sweet.home docker           --a-rec 192.168.2.249
+ipa dnsrecord-add sweet.home nix-cache        --a-rec 192.168.2.120
+ipa dnsrecord-add sweet.home pxe-boot         --a-rec 192.168.2.247
+ipa dnsrecord-add sweet.home tailscale-router --a-rec 192.168.2.121
+ipa dnsrecord-add sweet.home tor-relay        --a-rec 192.168.2.107
+ipa dnsrecord-add sweet.home pdm              --a-rec 192.168.2.248
+ipa dnsrecord-add sweet.home router           --a-rec 192.168.2.254
+```
+
+### 1c. Clean up stale reverse-zone PTR records
+
+FreeIPA already has PTR records from an earlier import but some are wrong.
+Fix them now so reverse DNS is accurate from day one.
+
+```bash
+# Remove stale "win11" entry at .250 (should be pve1)
+ipa dnsrecord-del 2.168.192.in-addr.arpa 250 --ptr-rec win11.
+ipa dnsrecord-add 2.168.192.in-addr.arpa 250 --ptr-rec pve1.sweet.home.
+
+# Fix unqualified PTR records (missing .sweet.home. suffix)
+ipa dnsrecord-mod 2.168.192.in-addr.arpa 108 --ptr-rec pbs.sweet.home.
+ipa dnsrecord-mod 2.168.192.in-addr.arpa 248 --ptr-rec pdm.sweet.home.
+ipa dnsrecord-mod 2.168.192.in-addr.arpa 249 --ptr-rec docker.sweet.home.
+ipa dnsrecord-mod 2.168.192.in-addr.arpa 252 --ptr-rec server.sweet.home.
+
+# Add any missing PTR records
+ipa dnsrecord-add 2.168.192.in-addr.arpa 119 --ptr-rec nixos.sweet.home.
+ipa dnsrecord-add 2.168.192.in-addr.arpa 120 --ptr-rec nix-cache.sweet.home.
+ipa dnsrecord-add 2.168.192.in-addr.arpa 121 --ptr-rec tailscale-router.sweet.home.
+ipa dnsrecord-add 2.168.192.in-addr.arpa 247 --ptr-rec pxe-boot.sweet.home.
+ipa dnsrecord-add 2.168.192.in-addr.arpa 254 --ptr-rec router.sweet.home.
+```
+
+### 1d. Point domain-controller's own DNS at itself
+
+```bash
+sudo nmcli connection modify "System eth0" ipv4.dns "127.0.0.1"
+sudo nmcli connection up "System eth0"
+```
+
+**Verify:**
+```bash
+dig pve1.sweet.home +short          # must return 192.168.2.250
+dig google.com +short               # must return an IP (NextDNS forwarding)
+```
+
+**Rollback 1d:** `sudo nmcli connection modify "System eth0" ipv4.dns "192.168.2.253" && sudo nmcli connection up "System eth0"`
+
+---
+
+## Stage 2 — Move pxe-boot DHCP options off Pi-hole (zero downtime)
+
+Pi-hole's dnsmasq currently serves the iPXE boot options via
+`99-ipxe-chainload.conf`. Before Pi-hole is retired, that config must move to
+the pxe-boot CT running dnsmasq in proxy mode so PXE boot keeps working.
+
+### 2a. Add dnsmasq proxy config to the pxe-boot NixOS module
+
+In `modules/build-types/pxe-boot.nix`, add:
+
+```nix
+services.dnsmasq = {
+  enable = true;
+  settings = {
+    # Proxy mode: respond only to PXE DHCP requests, leave normal leases to router
+    dhcp-range = [ "192.168.2.0,proxy" ];
+    # iPXE client detection
+    dhcp-match = [
+      "set:ipxe,175"
+      "set:efi64,option:client-arch,7"
+      "set:efi64,option:client-arch,9"
+    ];
+    dhcp-userclass = "set:ipxe,iPXE";
+    # Boot file selection
+    dhcp-boot = [
+      "tag:ipxe,tag:efi64,http://${vars.pxeServerIp}/boot.ipxe"
+      "tag:ipxe,http://${vars.pxeServerIp}/boot.ipxe"
+      "tag:efi64,ipxe.efi,,${vars.pxeServerIp}"
+      "undionly.kpxe,,${vars.pxeServerIp}"
+    ];
+  };
+};
+```
+
+### 2b. Rebuild and deploy the pxe-boot CT
+
+```bash
+# On pve1 — build the new tarball
+nix build .#lxc-pxe-boot.config.system.build.tarball
+
+# Verify dnsmasq starts correctly in the CT after deploy
+ssh nixos@192.168.2.247 systemctl status dnsmasq
+```
+
+### 2c. Remove the iPXE config from Pi-hole
+
+In the Pi-hole CT, remove `/etc/dnsmasq.d/99-ipxe-chainload.conf` and
+restart the FTL service:
+
+```bash
+ssh wayne@pve1.sweet.home \
+  "sudo pct exec 100 -- bash -c 'rm /etc/dnsmasq.d/99-ipxe-chainload.conf && systemctl restart pihole-FTL'"
+```
+
+**Verify:** PXE boot a test machine — it should still get an iPXE response and
+reach the boot menu.
+
+**Rollback 2c:** restore the file from the Pi-hole config backup at
+`/etc/pihole/config_backups/` and restart pihole-FTL.
+
+---
+
+## Stage 3 — DHCP migration: Pi-hole → router (brief maintenance window)
+
+**Do this in the evening.** Existing DHCP leases stay valid during the
+switchover so connected devices don't drop — only new lease requests fail
+during the gap, which is under 60 seconds if you follow the steps in order.
+
+The key: configure the router's DHCP DNS option to point at `.253` (Pi-hole's
+current IP). This way, all new leases issued by the router still get the same
+DNS server address — clients never need to change their DNS config. When Pi-hole
+is retired and the DC takes `.253` in Stage 4, `.253` just starts answering
+differently. No client reconfiguration.
+
+### 3a. Pre-configure router DHCP (do not enable yet)
+
+Log into `http://192.168.2.254`, find the DHCP settings and fill in — but
+leave DHCP **disabled** until step 3b:
+
+| Setting | Value |
+|---|---|
+| Start IP | 192.168.2.10 |
+| End IP | 192.168.2.59 |
+| Subnet mask | 255.255.255.0 |
+| Gateway | 192.168.2.254 |
+| Primary DNS | 192.168.2.253 |
+| Secondary DNS | *(leave blank)* |
+| Lease time | 24h |
+
+Save without enabling.
+
+### 3b. Switchover (do steps in quick succession)
+
+1. **Disable Pi-hole DHCP:** Pi-hole admin UI → Settings → DHCP → uncheck
+   "DHCP server enabled" → Save
+2. **Enable router DHCP** immediately after step 1
+
+### 3c. Verify router DHCP is working
+
+On a phone or laptop, disconnect from WiFi and reconnect (or run
+`sudo dhclient -r && sudo dhclient` on a Linux host):
+
+```bash
+ip addr show        # IP should be in 192.168.2.10–59 range
+dig google.com      # should resolve (Pi-hole DNS still running at .253)
+dig pve1.sweet.home # should resolve via FreeIPA at .138 (relayed via Pi-hole)
+```
+
+Wait 10–15 minutes for the most active devices to renew their leases. There's
+no need to wait for all leases to expire before proceeding.
+
+**Rollback 3b:** Re-enable Pi-hole DHCP. Disable router DHCP. Done — existing
+leases remain valid so most devices are unaffected.
+
+---
+
+## Stage 4 — Move domain-controller from .138 to .253
+
+Pi-hole lives at `.253`. The DC must take `.253` the moment Pi-hole stops so
+clients that still have `.253` as their DNS server don't notice the change.
+Script these commands in advance and run them in rapid succession.
+
+**Pre-stage: have this SSH command ready before running step 4a:**
+```bash
+ssh wayne@192.168.2.138 "
+  sudo nmcli connection modify 'System eth0' \
+    ipv4.addresses '192.168.2.253/24' \
+    ipv4.gateway '192.168.2.254' \
+    ipv4.dns '127.0.0.1' \
+    ipv4.method manual && \
+  sudo nmcli connection up 'System eth0'
+"
+```
+
+**Also update the Proxmox VM config to match (run from pve1):**
+```bash
+sudo qm set 108 \
+  --ipconfig0 ip=192.168.2.253/24,gw=192.168.2.254 \
+  --nameserver 192.168.2.253
+```
+
+### 4a. Stop Pi-hole
+
+```bash
+ssh wayne@pve1.sweet.home "sudo pct stop 100"
+```
+
+### 4b. Immediately: change DC's IP to .253
+
+Run the pre-staged SSH command from above. You have ~30 seconds before any
+client notices Pi-hole is gone. If SSH to `.138` refuses (the IP is already
+changing), open a Proxmox console to VM 108 and run the `nmcli` commands
+there.
+
+### 4c. Update Proxmox VM config
+
+Run the pre-staged `qm set 108` command from above.
+
+### 4d. Verify
+
+```bash
+ssh wayne@192.168.2.253   # must connect (new DC IP)
+dig @192.168.2.253 pve1.sweet.home +short   # must return 192.168.2.250
+dig @192.168.2.253 google.com +short        # must return an IP
+```
+
+From a client device that renewed its DHCP lease in Stage 3:
+```bash
+cat /etc/resolv.conf    # should show 192.168.2.253
+dig pve1.sweet.home     # should resolve
+```
+
+**Rollback 4:** `ssh wayne@pve1.sweet.home "sudo pct start 100"`. Change DC IP
+back to .138 via Proxmox console. This restores full Pi-hole DNS/DHCP service.
+Leave Pi-hole CT stopped-but-intact for 48 hours before deleting it.
+
+---
+
+## Stage 5 — Host renumbering (one at a time, any order)
+
+For each host:
+1. Update FreeIPA DNS A record and PTR record to the new IP
+2. Change the static IP on the host itself
+3. Verify SSH to new IP
+4. Update `variables.nix` if that host has an IP variable (pxe-boot, pbs — already done in this PR)
+
+**FreeIPA record update template** (run as admin on domain-controller):
+```bash
+ipa dnsrecord-mod sweet.home <hostname> --a-rec <new-ip>
+ipa dnsrecord-del 2.168.192.in-addr.arpa <old-last-octet> --ptr-rec <hostname>.sweet.home.
+ipa dnsrecord-add 2.168.192.in-addr.arpa <new-last-octet> --ptr-rec <hostname>.sweet.home.
+```
+
+### Renumbering order
+
+| # | Host | Old IP | New IP | How to change IP |
+|---|---|---|---|---|
+| 1 | nixos workstation | .119 | .243 | NetworkManager on guest; or `nmcli connection modify` |
+| 2 | nix-cache | .120 | .224 | `pct set 102 --net0 name=eth0,bridge=vmbr0,ip=192.168.2.224/24,gw=192.168.2.254` then `pct reboot 102` |
+| 3 | tailscale-router | .121 | .222 | Static config on guest; check Tailscale ACLs if IP is referenced there |
+| 4 | tor-relay | .107 | .221 | `pct set 104 --net0 name=eth0,bridge=vmbr0,ip=192.168.2.221/24,gw=192.168.2.254` then `pct reboot 104` |
+| 5 | pdm | .248 | .220 | `pct set 106 --net0 name=eth0,bridge=vmbr0,ip=192.168.2.220/24,gw=192.168.2.254` then `pct reboot 106` |
+| 6 | pxe-boot | .247 | .223 | `pct set 103 --net0 name=eth0,bridge=vmbr0,ip=192.168.2.223/24,gw=192.168.2.254` then rebuild NixOS (already updated in variables.nix) |
+| 7 | server | .252 | .226 | Static config on guest; NFS clients (docker) lose mounts briefly — they remount automatically |
+| 8 | docker | .249 | .225 | Static config on guest; do this after server is at .226 |
+| 9 | pbs | .108 | .244 | Static config on PBS host itself; update in `pbsIp` already done in variables.nix |
+| 10 | pve1 | .250 | .245 | Edit `/etc/network/interfaces` on the Proxmox host — see below |
+
+### pve1 renumber (step 10 — do last)
+
+All guests keep running; only the Proxmox web UI is briefly unreachable.
+
+```bash
+ssh wayne@pve1.sweet.home
+
+# Edit /etc/network/interfaces: change address from .250 to .245
+sudo nano /etc/network/interfaces
+# Change:  address 192.168.2.250/24
+# To:      address 192.168.2.245/24
+
+sudo systemctl restart networking
+# SSH will drop here — reconnect to new IP
+```
+
+```bash
+ssh wayne@192.168.2.245    # verify
+```
+
+Update FreeIPA DNS:
+```bash
+ipa dnsrecord-mod sweet.home pve1 --a-rec 192.168.2.245
+ipa dnsrecord-del 2.168.192.in-addr.arpa 250 --ptr-rec pve1.sweet.home.
+ipa dnsrecord-add 2.168.192.in-addr.arpa 245 --ptr-rec pve1.sweet.home.
+```
+
+**Rollback any step 5 host:** change the IP back on the guest and update the
+FreeIPA record back to the old IP. The old IP is unoccupied so you can
+temporarily use either.
+
+---
+
+## Stage 6 — Final cleanup
+
+Once all hosts are at their new IPs and verified:
+
+```bash
+# Delete the Pi-hole CT (already stopped since Stage 4)
+ssh wayne@pve1.sweet.home "sudo pct destroy 100"
+
+# Remove stale FreeIPA records for retired addresses
+ipa dnsrecord-del sweet.home pihole --del-all
+ipa dnsrecord-del 2.168.192.in-addr.arpa 253 --ptr-rec pihole.sweet.home.
+
+# Rebuild any NixOS hosts that reference pbsIp or pxeServerIp to pick up
+# the updated variables.nix values (pxe-boot mandatory; others as convenient)
+```
+
+---
+
+## Rollback summary
+
+| What broke | How to roll back |
+|---|---|
+| FreeIPA DNS not resolving | Check `systemctl status named` on DC; restart if failed |
+| FreeIPA DNS unreachable | `pct start 100` on pve1 (restores Pi-hole) |
+| Router DHCP not handing out leases | Re-enable Pi-hole DHCP; disable router DHCP |
+| DC unreachable after IP change | Proxmox console on VM 108 → `nmcli connection up "System eth0"` with old IP |
+| Host unreachable after renumber | Proxmox console → revert IP; or `pct set <id> --net0 ...` old IP and reboot CT |
+| pve1 web UI gone after renumber | SSH to .245 and check `/etc/network/interfaces`; if wrong, fix and restart networking |

docs/pxe-boot.md

@@ -15,6 +15,7 @@ rescue/inspection use.
 - TFTP root for first-stage bootloaders: `/srv/pxe/tftp`
 - iPXE entry script: `/srv/pxe/http/boot.ipxe`
 - Generated iPXE menu: `/srv/pxe/http/menu.ipxe`
+- Debian Minimal iPXE script: `/srv/pxe/http/debian.ipxe`
 - SystemRescue iPXE script: `/srv/pxe/http/systemrescue.ipxe`
 - TFTP fallback script: `/srv/pxe/tftp/autoexec.ipxe`
 - Boot binaries copied from the Nix `ipxe` package:
@@ -28,32 +29,51 @@ The host creates these directories with systemd tmpfiles:
 ```text
 /srv/pxe
 /srv/pxe/http
-/srv/pxe/http/images
+/srv/pxe/http/images  -> /mnt/pxe-images  (symlink to NFS share)
 /srv/pxe/http/auto-installer
 /srv/pxe/http/nixos-minimal
+/srv/pxe/http/debian
 /srv/pxe/http/systemrescue
 /srv/pxe/http/ubuntu
 /srv/pxe/http/rescue
 /srv/pxe/tftp
 ```
 
-Mount shared image storage under `/srv/pxe/http`, preferably
+`/srv/pxe/http/images` is a symlink to `/mnt/pxe-images`, which is an NFS
-`/srv/pxe/http/images` unless a menu entry expects files in a specific
+mount of `server.sweet.home:/tank/pxe-boot/images`
-directory such as `/srv/pxe/http/auto-installer`.
+(`modules/pxe-boot/mount-pxe-images.nix`). Place large images there (ISOs,
+disk images) rather than on the pxe-boot host's own root disk. For an LXC
+pxe-boot container the mount uses NFSv3+nolock with `nofail` (eager,
+non-blocking on server unavailability); for a Proxmox VM it uses NFSv4.2
+with `x-systemd.automount` (lazy, triggered on first access).
+
+When running as `lxc-pxe-boot`, the Proxmox container must have
+`features: nesting=1,mount=nfs` (at minimum) in its Proxmox config. `nesting=1`
+is required by systemd 260+ for credential isolation (user namespace creation
+and internal move-mounts); without it, AppArmor denies both, and every
+systemd service that uses `PrivateUsers`, `PrivateDevices`, or credential
+passing fails on boot. `mount=nfs` allows the NFSv3 mount. Both are set
+automatically by `scripts/proxmox/create-proxmox-resource.sh` (via
+`PROXMOX_DEFAULT_LXC_FEATURES` in `scripts/env.sh` which defaults to
+`nesting=1,keyctl=1,mount=nfs;nfs4`). If you ever change these features
+manually via `pct set`, be sure to include both — `pct set` replaces the
+entire features string, it does not append to it.
 
 The HTTP iPXE chain is:
 
 ```text
 undionly.kpxe or ipxe.efi
   -> autoexec.ipxe from the TFTP root, when iPXE requests it
-  -> http://192.168.2.247/boot.ipxe
+  -> http://192.168.2.223/boot.ipxe
-  -> http://192.168.2.247/menu.ipxe
+  -> http://192.168.2.223/menu.ipxe
 ```
 
 The generated menu currently exposes entries for:
 
 - NixOS Auto-Installer
 - NixOS Minimal
+- Debian Minimal
+- FreeIPA Server (Rocky Linux 9)
 - SystemRescue environment
 - iPXE shell
 - Reboot
@@ -84,19 +104,82 @@ directory name (`auto-installer` / `nixos-minimal`), so each one's
 generated system name (`nixos-system-<name>-*`) is self-describing rather
 than the nixpkgs default of `nixos-system-nixos-*` for both.
 
+The Debian Minimal entry chains `http://<pxeServerIp>/debian.ipxe`, which loads
+the Debian bookworm netboot kernel and initrd from `/srv/pxe/http/debian/`. The
+`fetch-debian-netboot.service` oneshot downloads these files from
+`deb.debian.org` on first boot (idempotent — skips if files are already
+present):
+
+```text
+/srv/pxe/http/debian/linux      (Debian bookworm netboot kernel)
+/srv/pxe/http/debian/initrd.gz  (Debian bookworm netboot initrd)
+```
+
+The service requires outbound internet access on the pxe-boot host. To
+re-download (e.g. after a Debian point release), delete the files and restart
+the service:
+
+```bash
+rm /srv/pxe/http/debian/linux /srv/pxe/http/debian/initrd.gz
+systemctl restart fetch-debian-netboot.service
+```
+
+To update to a different Debian release, change `debianRelease` in
+`modules/build-types/pxe-boot.nix` and redeploy.
+
+The **FreeIPA Server (Rocky Linux 9)** entry chains
+`http://<pxeServerIp>/rocky-freeipa.ipxe`, which boots the Rocky Linux 9
+Anaconda installer with a Kickstart file (`rocky-freeipa.ks`) hosted on the
+same server. The `fetch-rocky-pxeboot.service` oneshot downloads the pxeboot
+kernel and initrd from the Rocky Linux mirror on first boot (idempotent):
+
+```text
+/srv/pxe/http/rocky/vmlinuz     (Rocky Linux 9 Anaconda pxeboot kernel)
+/srv/pxe/http/rocky/initrd.img  (Rocky Linux 9 Anaconda pxeboot initrd)
+```
+
+The Kickstart file is generated from the NixOS module and staged at
+`/srv/pxe/http/rocky-freeipa.ks`. It performs a fully unattended install:
+
+1. Installs Rocky Linux 9 with `ipa-server` + `ipa-server-dns` packages
+2. Configures static IP `192.168.2.138`, hostname `domain-controller.sweet.home`
+3. Creates user `wayne` with the `adminSshKey` from `variables.nix`
+4. Generates random IPA passwords and writes them to `/root/ipa-credentials.txt`
+5. Creates a `freeipa-first-boot.service` oneshot that runs `ipa-server-install`
+   on first reboot (~20 minutes)
+
+After the install completes:
+- SSH in as `wayne@domain-controller` using the admin key
+- Monitor FreeIPA install progress: `sudo tail -f /root/freeipa-install.log`
+- Retrieve credentials: `sudo cat /root/ipa-credentials.txt` (save to password manager)
+- Configure Pi-hole: `server=/sweet.home/192.168.2.138` in dnsmasq
+
+To refresh the pxeboot files (e.g. after a Rocky point release):
+
+```bash
+rm /srv/pxe/http/rocky/vmlinuz /srv/pxe/http/rocky/initrd.img
+systemctl restart fetch-rocky-pxeboot.service
+```
+
+To update to a different Rocky release, change `rockyRelease` in
+`modules/build-types/pxe-boot.nix` and redeploy.
+
 The SystemRescue entry expects the source ISO at:
 
 ```text
 /srv/pxe/http/images/systemrescue.iso
 ```
 
+Since `/srv/pxe/http/images` is the NFS-backed symlink, place the ISO on the
+NFS share at `server.sweet.home:/tank/pxe-boot/images/systemrescue.iso`.
+
 The `stage-systemrescue.service` oneshot extracts that ISO into:
 
 ```text
 /srv/pxe/http/systemrescue
 ```
 
-The rescue menu entry then chains `http://192.168.2.247/systemrescue.ipxe`,
+The rescue menu entry then chains `http://192.168.2.223/systemrescue.ipxe`,
 which loads the SystemRescue kernel and initramfs from the extracted tree and
 uses `archiso_http_srv` to fetch the squashfs payload over HTTP.
 
@@ -113,6 +196,13 @@ After deployment by an operator, basic service checks are:
 ```bash
 curl http://pxe-boot/boot.ipxe
 curl http://pxe-boot/menu.ipxe
+curl http://pxe-boot/debian.ipxe
+curl -I http://pxe-boot/debian/linux
+curl -I http://pxe-boot/debian/initrd.gz
+curl http://pxe-boot/rocky-freeipa.ipxe
+curl http://pxe-boot/rocky-freeipa.ks
+curl -I http://pxe-boot/rocky/vmlinuz
+curl -I http://pxe-boot/rocky/initrd.img
 curl http://pxe-boot/systemrescue.ipxe
 curl -I http://pxe-boot/systemrescue/sysresccd/boot/x86_64/vmlinuz
 curl -I http://pxe-boot/systemrescue/sysresccd/boot/x86_64/sysresccd.img

flake.nix

@@ -130,6 +130,9 @@
         lxc-tailscale-router = mkTarget { platform = "lxc"; buildType = "tailscale-router"; hostPath = ./hosts/tailscale-router/host.nix; };
 
         lxc-tor-relay = mkTarget { platform = "lxc"; buildType = "tor-relay"; hostPath = ./hosts/tor-relay/host.nix; };
+
+        proxmox-ha-server-1 = mkTarget { platform = "proxmox"; buildType = "ha-server"; hostPath = ./hosts/ha-server-1/host.nix; };
+        proxmox-ha-server-2 = mkTarget { platform = "proxmox"; buildType = "ha-server"; hostPath = ./hosts/ha-server-2/host.nix; };
       };
 
       # Auto-install environments (migrated from the former nix-auto-installer

hosts/docker/host.nix

@@ -1,8 +1,17 @@
-_:
+{ vars, ... }:
 
 {
-  networking.hostName = "docker";
+  networking = {
-  networking.hostId = "007f0200";
+    hostName = "docker";
+    hostId = "007f0200";
+    useDHCP = false;
+    interfaces.${vars.vmLanInterface}.ipv4.addresses = [{
+      address = vars.dockerIp;
+      prefixLength = vars.lanPrefixLength;
+    }];
+    defaultGateway = { address = vars.lanGateway; interface = vars.vmLanInterface; };
+    nameservers = [ vars.domainControllerIp ];
+  };
   boot.zfs.forceImportRoot = false;
 
   # Preserved from the pre-refactor `docker` target — stateVersion must never

hosts/ha-server-1/host.nix

@@ -0,0 +1,30 @@
+{ vars, ... }:
+{
+  imports = [
+    (import ../../modules/beszel/host-token.nix {
+      name = "ha-server-1";
+      sopsFile = ../../secrets/ha-server-1.yaml;
+    })
+  ];
+
+  networking = {
+    hostName = vars.haServer1Host;
+    hostId = "3a4b5c6d";
+    useDHCP = false;
+    interfaces.${vars.vmLanInterface}.ipv4.addresses = [{
+      address = vars.haServer1Ip;
+      prefixLength = vars.lanPrefixLength;
+    }];
+    interfaces.${vars.vmStorageInterface}.ipv4.addresses = [{
+      address = vars.haServer1StorageIp;
+      prefixLength = vars.haStoragePrefixLength;
+    }];
+    defaultGateway = { address = vars.lanGateway; interface = vars.vmLanInterface; };
+    nameservers = [ vars.domainControllerIp ];
+  };
+
+  # Set KEY after pairing this host with the beszel hub; the token is sops-managed.
+  services.beszel.agent.environment.KEY = "";
+
+  system.stateVersion = "26.05";
+}

hosts/ha-server-2/host.nix

@@ -0,0 +1,30 @@
+{ vars, ... }:
+{
+  imports = [
+    (import ../../modules/beszel/host-token.nix {
+      name = "ha-server-2";
+      sopsFile = ../../secrets/ha-server-2.yaml;
+    })
+  ];
+
+  networking = {
+    hostName = vars.haServer2Host;
+    hostId = "7e8f9a0b";
+    useDHCP = false;
+    interfaces.${vars.vmLanInterface}.ipv4.addresses = [{
+      address = vars.haServer2Ip;
+      prefixLength = vars.lanPrefixLength;
+    }];
+    interfaces.${vars.vmStorageInterface}.ipv4.addresses = [{
+      address = vars.haServer2StorageIp;
+      prefixLength = vars.haStoragePrefixLength;
+    }];
+    defaultGateway = { address = vars.lanGateway; interface = vars.vmLanInterface; };
+    nameservers = [ vars.domainControllerIp ];
+  };
+
+  # Set KEY after pairing this host with the beszel hub; the token is sops-managed.
+  services.beszel.agent.environment.KEY = "";
+
+  system.stateVersion = "26.05";
+}

hosts/nix-cache/host.nix

@@ -8,7 +8,16 @@
     })
   ];
 
-  networking.hostName = vars.nixCacheHost;
+  networking = {
+    hostName = vars.nixCacheHost;
+    useDHCP = false;
+    interfaces.${vars.lxcLanInterface}.ipv4.addresses = [{
+      address = vars.nixCacheIp;
+      prefixLength = vars.lanPrefixLength;
+    }];
+    defaultGateway = { address = vars.lanGateway; interface = vars.lxcLanInterface; };
+    nameservers = [ vars.domainControllerIp ];
+  };
 
   services.beszel.agent.environment = {
     #DOCKER_HOST = "tcp://docker-socket-proxy:2375";

hosts/nixos/home.nix

@@ -27,7 +27,7 @@
     # Optional: set environment vars
     sessionVariables = {
       EDITOR = "vim";
-      SOPS_AGE_KEY_FILE = "~/.config/sops/age/keys.txt";
+      SOPS_AGE_KEY_FILE = "${config.home.homeDirectory}/.config/sops/age/keys.txt";
     };
 
     file = {

hosts/pxe-boot/host.nix

@@ -1,7 +1,16 @@
-_:
+{ vars, ... }:
 
 {
-  networking.hostName = "pxe-boot";
+  networking = {
+    hostName = "pxe-boot";
+    useDHCP = false;
+    interfaces.${vars.lxcLanInterface}.ipv4.addresses = [{
+      address = vars.pxeServerIp;
+      prefixLength = vars.lanPrefixLength;
+    }];
+    defaultGateway = { address = vars.lanGateway; interface = vars.lxcLanInterface; };
+    nameservers = [ vars.domainControllerIp ];
+  };
 
   # Preserved from the pre-refactor `pxe-boot` target — stateVersion must
   # never be bumped on an already-installed machine.

hosts/server/host.nix

@@ -8,8 +8,17 @@
     })
   ];
 
-  networking.hostName = vars.nfsServerHost;
+  networking = {
-  networking.hostId = "6689f93e";
+    hostName = vars.nfsServerHost;
+    hostId = "6689f93e";
+    useDHCP = false;
+    interfaces.${vars.vmLanInterface}.ipv4.addresses = [{
+      address = vars.serverIp;
+      prefixLength = vars.lanPrefixLength;
+    }];
+    defaultGateway = { address = vars.lanGateway; interface = vars.vmLanInterface; };
+    nameservers = [ vars.domainControllerIp ];
+  };
 
   services.beszel.agent.environment = {
     #DOCKER_HOST = "tcp://docker-socket-proxy:2375";

hosts/tailscale-router/host.nix

@@ -1,7 +1,27 @@
-_:
+{ vars, ... }:
 
 {
-  networking.hostName = "tailscale-router";
+  imports = [
+    (import ../../modules/beszel/host-token.nix {
+      name = "tailscale-router";
+      sopsFile = ../../secrets/tailscale-router.yaml;
+    })
+  ];
+
+  networking = {
+    hostName = "tailscale-router";
+    useDHCP = false;
+    interfaces.${vars.lxcLanInterface}.ipv4.addresses = [{
+      address = vars.tailscaleRouterIp;
+      prefixLength = vars.lanPrefixLength;
+    }];
+    defaultGateway = { address = vars.lanGateway; interface = vars.lxcLanInterface; };
+    nameservers = [ vars.domainControllerIp ];
+  };
+
+  services.beszel.agent.environment = {
+    KEY = "ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIFPR9kwtC4TAeTRu46A7+opZsYpxqkRJ+x/ZyB2GWCeG";
+  };
 
   # No networking.hostId: only ZFS-touching hosts (server, docker) need one
   # for pool-import safety, and this host does neither.

hosts/tor-relay/host.nix

@@ -1,4 +1,4 @@
-{ ... }:
+{ vars, ... }:
 
 {
   imports = [
@@ -8,10 +8,19 @@
     })
   ];
 
-  networking.hostName = "tor-relay";
+  networking = {
+    hostName = "tor-relay";
+    useDHCP = false;
+    interfaces.${vars.lxcLanInterface}.ipv4.addresses = [{
+      address = vars.torRelayIp;
+      prefixLength = vars.lanPrefixLength;
+    }];
+    defaultGateway = { address = vars.lanGateway; interface = vars.lxcLanInterface; };
+    nameservers = [ vars.domainControllerIp ];
+  };
 
-  # No networking.hostId: only ZFS-touching hosts (server, docker) need one
+  # No networking.hostId: only ZFS-touching hosts need one for pool-import
-  # for pool-import safety, and this host does neither.
+  # safety, and this host does neither.
 
   services.beszel.agent.environment = {
     KEY = "ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIFPR9kwtC4TAeTRu46A7+opZsYpxqkRJ+x/ZyB2GWCeG";

modules/build-types/docker.nix

@@ -15,7 +15,6 @@
     ../docker/enable-service.nix
     ../docker/nextcloud-cron-job.nix
     ../docker/docker-health-to-gotify.nix
-    ../tailscale/enable-service.nix
     ../traefik/rotate-logs.nix
     ../raspi/mount-data.nix
     ../services/enable-rpcbind.nix

modules/build-types/ha-server.nix

@@ -0,0 +1,43 @@
+# HA file server build type: DRBD + XFS + LIO iSCSI + NFS, managed by
+# Corosync + Pacemaker.  Both ha-server-1 and ha-server-2 use this type.
+#
+# NFS start/stop:
+#   services.nfs.server.enable = true configures /etc/exports, wires up
+#   rpcbind, and loads kernel modules — but nfs-server.service.wantedBy is
+#   force-cleared so systemd does NOT auto-start it at boot.  Pacemaker's
+#   ha-group resource group (configured by scripts/ha/cluster-init.sh)
+#   starts and stops nfs-server as part of the failover sequence after the
+#   XFS mount and iSCSI target are brought up on the new Active node.
+#
+# Beszel agent:
+#   Enabled here via enable-agent.nix.  The agent KEY (used to pair with
+#   the Beszel hub) is not set yet — add it to hosts/ha-server-{1,2}/host.nix
+#   under services.beszel.agent.environment.KEY once the hub accepts the
+#   new agents, following the pattern in hosts/server/host.nix.
+{ lib, vars, ... }:
+{
+  imports = [
+    ../ha/pacemaker-stack.nix
+    ../ha/iscsi-target.nix
+    ../ha/cluster-config.nix
+    ../beszel/enable-agent.nix
+  ];
+
+  services.nfs.server = {
+    enable = true;
+    exports = ''
+      ${vars.haStorageRoot}/${vars.nfsShares.dockerConfig.subpath} ${vars.lanCidr}${vars.nfsShares.options}
+      ${vars.haStorageRoot}/${vars.nfsShares.dockerVolumes.subpath} ${vars.lanCidr}${vars.nfsShares.options}
+      ${vars.haStorageRoot}/${vars.nfsShares.dockerDatabases.subpath} ${vars.lanCidr}${vars.nfsShares.options}
+      ${vars.haStorageRoot}/${vars.nfsShares.nextcloudData.subpath} ${vars.lanCidr}${vars.nfsShares.options}
+      ${vars.haStorageRoot}/${vars.nfsShares.raspiVolumes.subpath} ${vars.lanCidr}${vars.nfsShares.options}
+      ${vars.haStorageRoot}/${vars.nfsShares.proxmoxIsos.subpath} ${vars.lanCidr}${vars.nfsShares.options}
+      ${vars.haStorageRoot}/${vars.nfsShares.proxmoxLxcImages.subpath} ${vars.lanCidr}${vars.nfsShares.options}
+      ${vars.haStorageRoot}/${vars.nfsShares.pxebootImages.subpath} ${vars.lanCidr}${vars.nfsShares.options}
+    '';
+  };
+
+  # Pacemaker controls nfs-server — prevent systemd from starting it at boot
+  # on both nodes (only the Active node should be serving NFS).
+  systemd.services.nfs-server.wantedBy = lib.mkForce [ ];
+}

modules/build-types/pxe-boot.nix

@@ -21,6 +21,216 @@ let
     chain ${pxeBaseUrl}/boot.ipxe
   '';
 
+  debianRelease = "bookworm";
+  debianMirror = "https://deb.debian.org/debian";
+  debianNetbootBase = "${debianMirror}/dists/${debianRelease}/main/installer-amd64/current/images/netboot/debian-installer/amd64";
+
+  rockyRelease = "9";
+  rockyArch = "x86_64";
+  rockyMirror = "https://dl.rockylinux.org/pub/rocky/${rockyRelease}";
+  rockyPxebootBase = "${rockyMirror}/BaseOS/${rockyArch}/os/images/pxeboot";
+
+  debianIpxe = pkgs.writeText "debian.ipxe" ''
+    #!ipxe
+
+    set base ${pxeBaseUrl}
+
+    kernel ''${base}/debian/linux
+    initrd ''${base}/debian/initrd.gz
+    boot
+  '';
+
+  fetchDebianNetboot = pkgs.writeShellScript "fetch-debian-netboot" ''
+    set -eu
+
+    dir="${httpRoot}/debian"
+    mirror="${debianNetbootBase}"
+
+    if [ -f "$dir/linux" ] && [ -f "$dir/initrd.gz" ]; then
+      echo "Debian ${debianRelease} netboot files already present; skipping download."
+      exit 0
+    fi
+
+    echo "Downloading Debian ${debianRelease} netboot kernel and initrd from $mirror ..."
+    ${pkgs.curl}/bin/curl -fsSL -o "$dir/linux.tmp" "$mirror/linux"
+    ${pkgs.curl}/bin/curl -fsSL -o "$dir/initrd.gz.tmp" "$mirror/initrd.gz"
+    mv "$dir/linux.tmp" "$dir/linux"
+    mv "$dir/initrd.gz.tmp" "$dir/initrd.gz"
+    echo "Debian ${debianRelease} netboot files staged."
+  '';
+
+  # Rocky Linux 9 iPXE script — boots vmlinuz+initrd.img from the staged
+  # /rocky/ directory and hands Anaconda the hosted Kickstart URL.
+  # net.ifnames=0 biosdevname=0 ensures the NIC is eth0 in both the
+  # installer and the installed system (matches the Kickstart NM config).
+  rockyFreeIpaIpxe = pkgs.writeText "rocky-freeipa.ipxe" ''
+    #!ipxe
+
+    set base ${pxeBaseUrl}
+
+    kernel ''${base}/rocky/vmlinuz inst.ks=''${base}/rocky-freeipa.ks inst.repo=${rockyMirror}/BaseOS/${rockyArch}/os/ net.ifnames=0 biosdevname=0 ip=dhcp quiet
+    initrd ''${base}/rocky/initrd.img
+    boot
+  '';
+
+  # Kickstart file for domain-controller.sweet.home.
+  # Installs Rocky Linux 9, sets a static IP, creates wayne with the
+  # admin SSH key, then on first reboot runs ipa-server-install via a
+  # systemd oneshot service.  Passwords are generated at %post time,
+  # written to /root/ipa-credentials.txt (chmod 600), and read back by
+  # the first-boot script — never hardcoded here or in the repo.
+  rockyFreeIpaKs = pkgs.writeText "rocky-freeipa.ks" ''
+    #version=RHEL9
+    # Unattended Rocky Linux 9 + FreeIPA install
+    # Target: domain-controller.${vars.homeDomain}  ${vars.domainControllerIp}
+
+    url --url=${rockyMirror}/BaseOS/${rockyArch}/os/
+    repo --name=appstream --baseurl=${rockyMirror}/AppStream/${rockyArch}/os/
+
+    lang en_US.UTF-8
+    keyboard us
+    timezone UTC --utc
+
+    # DHCP during install; static IP configured in %post via NM config file
+    network --bootproto=dhcp --device=link --activate
+    network --hostname=domain-controller.sweet.home
+
+    selinux --enforcing
+    firewall --enabled --service=ssh
+
+    rootpw --lock
+    user --name=wayne --groups=wheel --shell=/bin/bash
+    sshkey --username=wayne "${vars.adminSshKey}"
+
+    zerombr
+    clearpart --all --initlabel --drives=sda
+    # Keep net.ifnames=0 biosdevname=0 in the installed GRUB so the NIC
+    # stays eth0 after reboot (matches the NM connection file below).
+    bootloader --location=mbr --boot-drive=sda --append="net.ifnames=0 biosdevname=0"
+
+    part /boot --fstype=xfs --size=1024 --ondisk=sda
+    part swap  --fstype=swap --size=2048 --ondisk=sda
+    part /     --fstype=xfs  --grow --size=1 --ondisk=sda --asprimary
+
+    %packages
+    @^minimal-environment
+    ipa-server
+    ipa-server-dns
+    %end
+
+    reboot
+
+    %post --log=/root/ks-post.log
+    set -euo pipefail
+
+    # -- Static IP: write NM connection file directly (NM not running in chroot) --
+    mkdir -p /etc/NetworkManager/system-connections
+    cat > /etc/NetworkManager/system-connections/eth0.nmconnection << 'NMCONN'
+    [connection]
+    id=eth0
+    type=ethernet
+    interface-name=eth0
+    autoconnect=true
+
+    [ethernet]
+
+    [ipv4]
+    method=manual
+    addresses=${vars.domainControllerIp}/${toString vars.lanPrefixLength}
+    gateway=${vars.lanGateway}
+    dns=${vars.domainControllerIp};
+    dns-search=${vars.homeDomain};
+
+    [ipv6]
+    method=auto
+    NMCONN
+    chmod 600 /etc/NetworkManager/system-connections/eth0.nmconnection
+
+    # -- /etc/hosts: FQDN must resolve to the real IP (not loopback) for IPA --
+    sed -i '/domain-controller/d' /etc/hosts
+    echo '${vars.domainControllerIp} domain-controller.${vars.homeDomain} domain-controller' >> /etc/hosts
+
+    # -- Generate IPA passwords and store securely --
+    DM_PASS=$(openssl rand -base64 24 | tr -dc 'A-Za-z0-9' | head -c 24)
+    ADMIN_PASS=$(openssl rand -base64 24 | tr -dc 'A-Za-z0-9' | head -c 24)
+    printf 'Directory Manager: %s\nIPA Admin: %s\n' "$DM_PASS" "$ADMIN_PASS" \
+      > /root/ipa-credentials.txt
+    chmod 600 /root/ipa-credentials.txt
+
+    # -- First-boot script: reads passwords back, runs ipa-server-install --
+    cat > /usr/local/sbin/freeipa-first-boot.sh << 'FIRSTBOOT'
+    #!/bin/bash
+    set -euo pipefail
+    exec >> /root/freeipa-install.log 2>&1
+    echo "=== FreeIPA first-boot install started at $(date) ==="
+
+    DM_PASS=$(grep '^Directory Manager:' /root/ipa-credentials.txt | awk '{print $NF}')
+    ADMIN_PASS=$(grep '^IPA Admin:' /root/ipa-credentials.txt | awk '{print $NF}')
+
+    ipa-server-install \
+      --realm=SWEET.HOME \
+      --domain=sweet.home \
+      --hostname=domain-controller.sweet.home \
+      --ds-password="$DM_PASS" \
+      --admin-password="$ADMIN_PASS" \
+      --setup-dns \
+      --forwarder=192.168.2.253 \
+      --no-dnssec-validation \
+      --no-ntp \
+      --unattended
+
+    echo "=== FreeIPA install complete at $(date) ==="
+    echo "Credentials: /root/ipa-credentials.txt (save to password manager)"
+    echo "CA backup:   /root/cacert.p12 (encrypted with Directory Manager password)"
+    systemctl disable freeipa-first-boot.service
+    FIRSTBOOT
+    chmod 700 /usr/local/sbin/freeipa-first-boot.sh
+
+    # -- Systemd oneshot service: runs freeipa-first-boot.sh on first real boot --
+    cat > /etc/systemd/system/freeipa-first-boot.service << 'UNIT'
+    [Unit]
+    Description=FreeIPA first-boot installation
+    After=network-online.target
+    Wants=network-online.target
+    ConditionPathExists=/root/ipa-credentials.txt
+
+    [Service]
+    Type=oneshot
+    ExecStart=/usr/local/sbin/freeipa-first-boot.sh
+    TimeoutStartSec=1800
+    RemainAfterExit=yes
+
+    [Install]
+    WantedBy=multi-user.target
+    UNIT
+
+    mkdir -p /etc/systemd/system/multi-user.target.wants
+    ln -sf /etc/systemd/system/freeipa-first-boot.service \
+      /etc/systemd/system/multi-user.target.wants/freeipa-first-boot.service
+
+    echo "Kickstart %post complete. FreeIPA installs on first reboot (~20 min)."
+    %end
+  '';
+
+  fetchRockyPxeboot = pkgs.writeShellScript "fetch-rocky-pxeboot" ''
+    set -eu
+
+    dir="${httpRoot}/rocky"
+    base="${rockyPxebootBase}"
+
+    if [ -f "$dir/vmlinuz" ] && [ -f "$dir/initrd.img" ]; then
+      echo "Rocky Linux ${rockyRelease} pxeboot files already present; skipping download."
+      exit 0
+    fi
+
+    echo "Downloading Rocky Linux ${rockyRelease} pxeboot kernel and initrd from $base ..."
+    ${pkgs.curl}/bin/curl -fsSL -o "$dir/vmlinuz.tmp"   "$base/vmlinuz"
+    ${pkgs.curl}/bin/curl -fsSL -o "$dir/initrd.img.tmp" "$base/initrd.img"
+    mv "$dir/vmlinuz.tmp"   "$dir/vmlinuz"
+    mv "$dir/initrd.img.tmp" "$dir/initrd.img"
+    echo "Rocky Linux ${rockyRelease} pxeboot files staged."
+  '';
+
   systemRescueIpxe = pkgs.writeText "systemrescue.ipxe" ''
     #!ipxe
 
@@ -68,11 +278,13 @@ let
     set base ${pxeBaseUrl}
 
     menu PXE Boot Menu
-    item auto-installer NixOS Auto-Installer
+    item auto-installer    NixOS Auto-Installer
-    item nixos-minimal NixOS Minimal
+    item nixos-minimal     NixOS Minimal
-    item rescue Rescue Environment
+    item debian            Debian Minimal
-    item shell iPXE Shell
+    item rocky-freeipa     FreeIPA Server (Rocky Linux 9)
-    item reboot Reboot
+    item rescue            Rescue Environment
+    item shell             iPXE Shell
+    item reboot            Reboot
 
     choose target && goto ''${target}
 
@@ -82,6 +294,12 @@ let
     :nixos-minimal
     chain ''${base}/nixos-minimal/netboot.ipxe
 
+    :debian
+    chain ''${base}/debian.ipxe
+
+    :rocky-freeipa
+    chain ''${base}/rocky-freeipa.ipxe
+
     :rescue
     chain ''${base}/systemrescue.ipxe
 
@@ -95,6 +313,7 @@ in
 {
   imports = [
     ../pxe-boot/stage-installer-artifacts.nix
+    ../pxe-boot/mount-pxe-images.nix
   ];
 
   environment.systemPackages = with pkgs; [
@@ -129,37 +348,101 @@ in
     openssh.settings.PermitRootLogin = "yes";
   };
 
-  systemd.tmpfiles.rules = [
+  systemd = {
-    "d ${pxeRoot} 0755 root root -"
+    tmpfiles.rules = [
-    "d ${httpRoot} 0755 root root -"
+      "d ${pxeRoot} 0755 root root -"
-    "d ${httpRoot}/images 0755 root root -"
+      "d ${httpRoot} 0755 root root -"
-    "d ${httpRoot}/auto-installer 0755 root root -"
+      "L+ ${httpRoot}/images - - - - ${vars.nfsShares.pxebootImages.mountpoint}"
-    "d ${httpRoot}/nixos-minimal 0755 root root -"
+      "d ${httpRoot}/auto-installer 0755 root root -"
-    "d ${httpRoot}/systemrescue 0755 root root -"
+      "d ${httpRoot}/nixos-minimal 0755 root root -"
-    "d ${httpRoot}/ubuntu 0755 root root -"
+      "d ${httpRoot}/systemrescue 0755 root root -"
-    "d ${httpRoot}/rescue 0755 root root -"
+      "d ${httpRoot}/debian 0755 root root -"
-    "d ${tftpRoot} 0755 root root -"
+      "d ${httpRoot}/ubuntu 0755 root root -"
-    "C+ ${httpRoot}/boot.ipxe 0644 root root - ${bootIpxe}"
+      "d ${httpRoot}/rescue 0755 root root -"
-    "C+ ${httpRoot}/menu.ipxe 0644 root root - ${menuIpxe}"
+      "d ${httpRoot}/rocky 0755 root root -"
-    "C+ ${httpRoot}/systemrescue.ipxe 0644 root root - ${systemRescueIpxe}"
+      "d ${tftpRoot} 0755 root root -"
-    "C+ ${tftpRoot}/autoexec.ipxe 0644 root root - ${autoexecIpxe}"
+      "C+ ${httpRoot}/boot.ipxe 0644 root root - ${bootIpxe}"
-    "C+ ${tftpRoot}/ipxe.efi 0644 root root - ${pkgs.ipxe}/ipxe.efi"
+      "C+ ${httpRoot}/menu.ipxe 0644 root root - ${menuIpxe}"
-    "C+ ${tftpRoot}/undionly.kpxe 0644 root root - ${pkgs.ipxe}/undionly.kpxe"
+      "C+ ${httpRoot}/debian.ipxe 0644 root root - ${debianIpxe}"
-  ];
+      "C+ ${httpRoot}/rocky-freeipa.ipxe 0644 root root - ${rockyFreeIpaIpxe}"
-
+      "C+ ${httpRoot}/rocky-freeipa.ks 0644 root root - ${rockyFreeIpaKs}"
-  systemd.services.stage-systemrescue = {
+      "C+ ${httpRoot}/systemrescue.ipxe 0644 root root - ${systemRescueIpxe}"
-    description = "Stage SystemRescue ISO contents for HTTP PXE boot";
+      "C+ ${tftpRoot}/autoexec.ipxe 0644 root root - ${autoexecIpxe}"
-    after = [
+      "C+ ${tftpRoot}/ipxe.efi 0644 root root - ${pkgs.ipxe}/ipxe.efi"
-      "local-fs.target"
+      "C+ ${tftpRoot}/undionly.kpxe 0644 root root - ${pkgs.ipxe}/undionly.kpxe"
-      "systemd-tmpfiles-setup.service"
     ];
-    wantedBy = [ "multi-user.target" ];
+
-    serviceConfig = {
+    services = {
-      Type = "oneshot";
+      fetch-debian-netboot = {
-      ExecStart = stageSystemRescue;
+        description = "Download Debian ${debianRelease} netboot kernel and initrd for HTTP PXE boot";
+        after = [
+          "local-fs.target"
+          "systemd-tmpfiles-setup.service"
+          "network-online.target"
+        ];
+        wants = [ "network-online.target" ];
+        wantedBy = [ "multi-user.target" ];
+        serviceConfig = {
+          Type = "oneshot";
+          ExecStart = fetchDebianNetboot;
+          RemainAfterExit = true;
+        };
+      };
+
+      fetch-rocky-pxeboot = {
+        description = "Download Rocky Linux ${rockyRelease} pxeboot kernel and initrd for HTTP PXE boot";
+        after = [
+          "local-fs.target"
+          "systemd-tmpfiles-setup.service"
+          "network-online.target"
+        ];
+        wants = [ "network-online.target" ];
+        wantedBy = [ "multi-user.target" ];
+        serviceConfig = {
+          Type = "oneshot";
+          ExecStart = fetchRockyPxeboot;
+          RemainAfterExit = true;
+        };
+      };
+
+      stage-systemrescue = {
+        description = "Stage SystemRescue ISO contents for HTTP PXE boot";
+        after = [
+          "local-fs.target"
+          "systemd-tmpfiles-setup.service"
+        ];
+        wantedBy = [ "multi-user.target" ];
+        serviceConfig = {
+          Type = "oneshot";
+          ExecStart = stageSystemRescue;
+        };
+      };
+    };
+  };
+
+  services.dnsmasq = {
+    enable = true;
+    settings = {
+      # Disable DNS listener — only proxy DHCP is needed here.
+      # Without this dnsmasq tries to bind port 53 which systemd-resolved
+      # already owns, causing startup failure.
+      port = 0;
+      dhcp-range = [ "192.168.2.0,proxy" ];
+      dhcp-match = [
+        "set:ipxe,175"
+        "set:efi64,option:client-arch,7"
+        "set:efi64,option:client-arch,9"
+      ];
+      dhcp-userclass = "set:ipxe,iPXE";
+      dhcp-boot = [
+        "tag:ipxe,tag:efi64,http://${vars.pxeServerIp}/boot.ipxe"
+        "tag:ipxe,http://${vars.pxeServerIp}/boot.ipxe"
+        "tag:efi64,ipxe.efi,,${vars.pxeServerIp}"
+        "undionly.kpxe,,${vars.pxeServerIp}"
+      ];
     };
   };
 
   networking.firewall.allowedTCPPorts = [ vars.ports.pxeBootHttp ];
-  networking.firewall.allowedUDPPorts = [ vars.ports.pxeBootTftp ];
+  networking.firewall.allowedUDPPorts = [ vars.ports.pxeBootTftp 67 ];
 }

modules/build-types/server.nix

@@ -10,7 +10,8 @@ let
     in lib.imap1 (i: _: lib.concatStringsSep "/" (lib.take i parts)) parts;
 
   poolDatasets = lib.unique (
-    lib.concatMap (share: ancestors share.subpath) (lib.attrValues vars.nfsShares)
+    lib.concatMap (share: ancestors share.subpath)
+      (lib.filter builtins.isAttrs (lib.attrValues vars.nfsShares))
   );
 in
 {
@@ -43,14 +44,8 @@ in
         exit 0
       fi
 
-      # Pool exists on a device but not yet imported — let the standard
+      # Locate the data disk first — used for both the fallback import
-      # zfs-import-${poolName}.service handle it normally.
+      # attempt and, only if the disk is genuinely blank, pool creation.
-      if zpool import -d /dev/disk/by-id -N "${poolName}" 2>/dev/null; then
-        exit 0
-      fi
-
-      # No pool found at all.  Create it on the Proxmox data disk (scsi1),
-      # which appears as /dev/disk/by-id/scsi-*drive-scsi1 inside the VM.
       DATA_DISK=""
       for candidate in /dev/disk/by-id/scsi-*drive-scsi1; do
         [[ "$candidate" == *-part* ]] && continue
@@ -62,8 +57,31 @@ in
         exit 1
       fi
 
+      # Try importing via the by-id symlink directory first (normal path),
+      # then fall back to scanning the disk directly.  The two-step exists
+      # because of a udev race: systemd-udev-settle.service can clear before
+      # /dev/disk/by-id/ entries are fully populated, causing the first
+      # import to fail even when the pool is intact on the disk.
+      if zpool import -d /dev/disk/by-id -N "${poolName}" 2>/dev/null; then
+        exit 0
+      fi
+      if zpool import -d "$DATA_DISK" -N "${poolName}" 2>/dev/null; then
+        exit 0
+      fi
+
+      # Both import attempts failed.  Before creating a new pool, verify the
+      # disk is genuinely blank — if ZFS label metadata is present the import
+      # failed for some other reason and we must not clobber existing data.
+      if zdb -l "$DATA_DISK" 2>/dev/null | grep -q "name: '${poolName}'"; then
+        echo "zfs-init-${poolName}: $DATA_DISK has ZFS pool '${poolName}' metadata but import failed — refusing to overwrite existing data. Run 'zpool import -d $DATA_DISK ${poolName}' manually to investigate." >&2
+        exit 1
+      fi
+
+      # Disk is genuinely blank: create the pool.  -f is intentionally
+      # omitted so that if we somehow reach this point with an existing pool
+      # on the disk, zpool refuses rather than silently destroying data.
       echo "zfs-init-${poolName}: creating pool on $DATA_DISK"
-      zpool create -f "${poolName}" "$DATA_DISK"
+      zpool create "${poolName}" "$DATA_DISK"
       ${lib.concatMapStrings (ds: ''
         zfs create "${poolName}/${ds}"
       '') poolDatasets}
@@ -78,13 +96,20 @@ in
   services.nfs.server = {
     enable = true;
     exports = ''
-      ${vars.storageRoot}/${vars.nfsShares.dockerConfig.subpath} ${vars.lanCidr}(rw,sync,no_subtree_check,no_root_squash)
+      ${vars.storageRoot}/${vars.nfsShares.dockerConfig.subpath} ${vars.lanCidr}${vars.nfsShares.options}
-      ${vars.storageRoot}/${vars.nfsShares.dockerVolumes.subpath} ${vars.lanCidr}(rw,sync,no_subtree_check,no_root_squash)
+      ${vars.storageRoot}/${vars.nfsShares.dockerVolumes.subpath} ${vars.lanCidr}${vars.nfsShares.options}
-      ${vars.storageRoot}/${vars.nfsShares.dockerDatabases.subpath} ${vars.lanCidr}(rw,sync,no_subtree_check,no_root_squash)
+      ${vars.storageRoot}/${vars.nfsShares.dockerDatabases.subpath} ${vars.lanCidr}${vars.nfsShares.options}
-      ${vars.storageRoot}/${vars.nfsShares.nextcloudData.subpath} ${vars.lanCidr}(rw,sync,no_subtree_check,no_root_squash)
+      ${vars.storageRoot}/${vars.nfsShares.nextcloudData.subpath} ${vars.lanCidr}${vars.nfsShares.options}
-      ${vars.storageRoot}/${vars.nfsShares.raspiVolumes.subpath} ${vars.lanCidr}(rw,sync,no_subtree_check,no_root_squash)
+      ${vars.storageRoot}/${vars.nfsShares.raspiVolumes.subpath} ${vars.lanCidr}${vars.nfsShares.options}
+      ${vars.storageRoot}/${vars.nfsShares.proxmoxIsos.subpath} ${vars.lanCidr}${vars.nfsShares.options}
+      ${vars.storageRoot}/${vars.nfsShares.proxmoxLxcImages.subpath} ${vars.lanCidr}${vars.nfsShares.options}
+      ${vars.storageRoot}/${vars.nfsShares.pxebootImages.subpath} ${vars.lanCidr}${vars.nfsShares.options}
     '';
   };
 
-  networking.firewall.allowedTCPPorts = [ vars.ports.nfsRpcbind vars.ports.nfsd ];
+  # mountd (20048) is needed for showmount/NFSv3 mount protocol — without it
+  # clients can reach portmapper (111) and get the mountd port back, then
+  # time out trying to connect to it. All three ports need TCP and UDP.
+  networking.firewall.allowedTCPPorts = [ vars.ports.nfsRpcbind vars.ports.nfsd vars.ports.nfsMountd ];
+  networking.firewall.allowedUDPPorts = [ vars.ports.nfsRpcbind vars.ports.nfsd vars.ports.nfsMountd ];
 }

modules/build-types/tailscale-router.nix

@@ -1,8 +1,10 @@
-{ ... }:
+{ vars, ... }:
 
 {
   imports = [
     ../tailscale/subnet-router.nix
+    ../tailscale/ts-dns-forwarder.nix
+    ../beszel/enable-agent.nix
   ];
 
   # "server", not "both": this build type advertises LAN subnet routes but
@@ -12,8 +14,31 @@
   # own setting) so the intent is clear at the build-type level.
   services.tailscale.useRoutingFeatures = "server";
 
-  # Forwarded subnet-router traffic arrives on tailscale0 already
+  # Advertise the LAN subnet so Tailscale peers can route back to LAN machines.
-  # tailscale-authenticated -- the firewall's normal per-port allow-list
+  # Must also be approved in the Tailscale admin console (Machines → Edit route settings).
-  # would otherwise drop it. Standard NixOS/Tailscale subnet-router guidance.
+  services.tailscale.extraUpFlags = [ "--advertise-routes=${vars.lanCidr}" ];
-  networking.firewall.trustedInterfaces = [ "tailscale0" ];
+
+  networking.firewall = {
+    # Forwarded subnet-router traffic arrives on tailscale0 already
+    # tailscale-authenticated -- the firewall's normal per-port allow-list
+    # would otherwise drop it. Standard NixOS/Tailscale subnet-router guidance.
+    trustedInterfaces = [ "tailscale0" ];
+
+    # SNAT LAN traffic going into Tailscale so the remote peer sees it as
+    # coming from this router's Tailscale IP rather than a raw LAN IP.
+    # Without this, Tailscale drops forwarded packets whose source is not a
+    # recognised Tailscale address.
+    #
+    # We target POSTROUTING directly (always-existing built-in chain) rather
+    # than nixos-nat-post: extraCommands runs after the old nixos-nat-post is
+    # deleted but before the new one is created, so -A nixos-nat-post silently
+    # fails.  The -C check makes the rule idempotent across firewall reloads.
+    extraCommands = ''
+      iptables -t nat -C POSTROUTING -s ${vars.lanCidr} -o tailscale0 -j MASQUERADE 2>/dev/null || \
+        iptables -t nat -A POSTROUTING -s ${vars.lanCidr} -o tailscale0 -j MASQUERADE
+    '';
+    extraStopCommands = ''
+      iptables -t nat -D POSTROUTING -s ${vars.lanCidr} -o tailscale0 -j MASQUERADE 2>/dev/null || true
+    '';
+  };
 }

modules/ha/cluster-config.nix

@@ -0,0 +1,107 @@
+# Cluster-wide HA config shared by both ha-server nodes.
+#
+# Covers everything that is identical on both nodes and references cluster
+# topology (node IPs, hostnames, DRBD resource).  Per-node identity
+# (hostname, static IP, stateVersion) lives in hosts/ha-server-{1,2}/host.nix.
+#
+# Corosync authkey:
+#   /etc/corosync/authkey (mode 0400) is managed by sops-nix below.
+#   Bootstrap: run scripts/ha/cluster-init.sh on node1 to generate the key,
+#   then encrypt it with: sops -e --input-type binary /etc/corosync/authkey > secrets/ha-corosync-authkey
+#   Both host keys must be registered via sync-host-keys.sh first so both nodes can decrypt it.
+#
+# DRBD fencing:
+#   Production setting is resource-only: DRBD waits for the STONITH fence
+#   agent to confirm the peer is dead before promoting to Primary.  This
+#   requires a working fence_pve_ssh STONITH resource in Pacemaker
+#   (see scripts/ha/cluster-enable-stonith.sh).  On a fresh cluster with
+#   no fence device yet, temporarily change to dont-care and run
+#   cluster-enable-stonith.sh once the fence key is deployed.
+{ lib, vars, ... }:
+{
+  services.drbd = {
+    enable = true;
+    config = ''
+      global {
+        usage-count yes;
+      }
+
+      common {
+        net {
+          protocol C;
+          ping-int 1;
+          verify-alg sha256;
+          after-sb-0pri discard-zero-changes;
+          after-sb-1pri discard-secondary;
+        }
+        disk {
+          fencing resource-only;
+        }
+      }
+
+      resource ha-data {
+        volume 0 {
+          device    /dev/drbd0;
+          disk      /dev/sdb;
+          meta-disk internal;
+        }
+
+        on ${vars.haServer1Host} {
+          address ${vars.haServer1StorageIp}:${toString vars.ports.haServerDrbd};
+        }
+
+        on ${vars.haServer2Host} {
+          address ${vars.haServer2StorageIp}:${toString vars.ports.haServerDrbd};
+        }
+      }
+    '';
+  };
+
+  # /etc/corosync/authkey — sops binary secret, identical on both nodes.
+  # Decryptable by both ha-server host keys (added by sync-host-keys.sh).
+  sops.secrets.corosync_authkey = {
+    sopsFile = ../../secrets/ha-corosync-authkey;
+    format = "binary";
+    path = "/etc/corosync/authkey";
+    mode = "0400";
+    restartUnits = [ "corosync.service" ];
+  };
+
+  # NixOS common config enables NetworkManager by default; HA cluster nodes
+  # need stable static IPs with predictable interface names — NM is not suitable.
+  networking.networkmanager.enable = lib.mkForce false;
+
+  # services.corosync.enable is set by modules/ha/pacemaker-stack.nix.
+  services.corosync = {
+    clusterName = "ha-cluster";
+    nodelist = [
+      { nodeid = 1; name = vars.haServer1Host; ring_addrs = [ vars.haServer1StorageIp ]; }
+      { nodeid = 2; name = vars.haServer2Host; ring_addrs = [ vars.haServer2StorageIp ]; }
+    ];
+  };
+
+  networking.firewall = {
+    allowedTCPPorts = [
+      vars.ports.haServerIscsi
+      vars.ports.haServerPacemakerRemoted
+      vars.ports.haServerPcsd
+      vars.ports.haServerDrbd
+      vars.ports.nfsRpcbind
+      vars.ports.nfsd
+      vars.ports.nfsMountd
+    ];
+    allowedUDPPorts = [
+      vars.ports.haServerCorosync1
+      vars.ports.haServerCorosync2
+      vars.ports.haServerCorosyncCrypto
+      vars.ports.nfsRpcbind
+      vars.ports.nfsd
+      vars.ports.nfsMountd
+    ];
+    extraCommands = ''
+      iptables -A INPUT -s ${vars.haServer1Ip}/32 -j ACCEPT
+      iptables -A INPUT -s ${vars.haServer2Ip}/32 -j ACCEPT
+      iptables -A INPUT -s ${vars.haStorageCidr} -j ACCEPT
+    '';
+  };
+}

modules/ha/iscsi-target.nix

@@ -0,0 +1,99 @@
+# LIO iSCSI target service (targetctl) for NixOS HA clusters.
+#
+# Provides the targetctl.service that saves/restores LIO configuration from
+# /etc/target/saveconfig.json.  Pacemaker manages this service via its
+# systemd resource agent (class="systemd" type="targetctl").
+#
+# Why ExecStop is not simply "targetctl save":
+#   targetctl save writes the LIO config to JSON but does NOT remove the LIO
+#   target from the kernel's configfs.  As a result, any fileio backing store
+#   that LIO has open (e.g. iscsi-lun.img on an XFS-over-DRBD filesystem)
+#   stays referenced in the kernel.  The subsequent XFS umount from the
+#   Filesystem OCF resource then returns EBUSY and either hangs for the full
+#   op-stop timeout or fails outright, blocking the entire failover.
+#
+#   The ExecStop script here additionally tears down the kernel LIO state
+#   via rtslib_fb after saving, so the backing-store file descriptor is
+#   released and umount succeeds immediately.
+#
+# Empty-config guard:
+#   The save step is skipped when no iSCSI targets are currently active.
+#   This prevents the secondary node (where LIO was never started) from
+#   overwriting a valid saveconfig.json with an empty one when Pacemaker
+#   stops the iscsi-target resource as part of a failover or cleanup.
+{ pkgs, ... }:
+
+let
+  python3 = pkgs.python3.withPackages (ps: [ ps.rtslib-fb ]);
+  targetctl = "${pkgs.targetcli-fb}/bin/targetctl";
+
+  targetctlStop = pkgs.writeScript "targetctl-stop" ''
+    #!${python3}/bin/python3
+    import subprocess, sys
+    import rtslib_fb
+
+    root = rtslib_fb.RTSRoot()
+    targets = list(root.targets)
+    if targets:
+        subprocess.run(
+            ["${targetctl}", "save", "/etc/target/saveconfig.json"],
+            capture_output=True,
+        )
+        print(f"saved {len(targets)} iSCSI target(s)")
+    else:
+        print("no active LIO targets — saveconfig.json unchanged")
+
+    for target in targets:
+        try:
+            for tpg in list(target.tpgs):
+                tpg.enable = False
+            target.delete()
+        except Exception as e:
+            print(f"warn (target): {e}", file=sys.stderr)
+    for so in list(root.storage_objects):
+        try:
+            so.delete()
+        except Exception as e:
+            print(f"warn (backstore): {e}", file=sys.stderr)
+    print("LIO kernel target cleared")
+  '';
+in
+{
+  boot.kernelModules = [
+    "target_core_mod"
+    "iscsi_target_mod"
+    "target_core_file"
+    "target_core_pscsi"
+    "target_core_user"
+    "configfs"
+  ];
+
+  systemd = {
+    mounts = [{
+      where = "/sys/kernel/config";
+      what = "configfs";
+      type = "configfs";
+      wantedBy = [ "multi-user.target" ];
+      before = [ "targetctl.service" ];
+    }];
+    services.targetctl = {
+      description = "LIO iSCSI target config save/restore";
+      wantedBy = [ "multi-user.target" ];
+      after = [ "sys-kernel-config.mount" "network.target" ];
+      requires = [ "sys-kernel-config.mount" ];
+      serviceConfig = {
+        Type = "oneshot";
+        RemainAfterExit = true;
+        ExecStart = "${targetctl} restore /etc/target/saveconfig.json";
+        ExecStop = "${targetctlStop}";
+      };
+      unitConfig.ConditionFileNotEmpty = "/etc/target/saveconfig.json";
+    };
+    tmpfiles.rules = [
+      "d /etc/target 0750 root root -"
+      "f /etc/target/saveconfig.json 0640 root root -"
+    ];
+  };
+
+  environment.systemPackages = [ pkgs.targetcli-fb ];
+}

modules/ha/pacemaker-stack.nix

@@ -0,0 +1,94 @@
+# Pacemaker + Corosync HA stack for NixOS with known-good workarounds.
+#
+# Issues fixed here (confirmed through live testing on NixOS 25.11):
+#
+# 1. StateDirectory ownership reset: systemd's StateDirectory=pacemaker
+#    creates /var/lib/pacemaker owned root:root.  pacemaker-based (the CIB
+#    daemon) runs as the hacluster user and calls pcmk__daemon_can_write,
+#    which requires the CIB directory to be owned by hacluster or be
+#    group-writable by haclient.  Workaround: remove StateDirectory and let
+#    ExecStartPre create every required subdirectory with correct ownership.
+#
+# 2. HA_SBIN_DIR wrong path: ocf-shellfuncs sets HA_SBIN_DIR to the Nix
+#    store path of the resource-agents derivation's /sbin, which doesn't
+#    exist.  The DRBD OCF agent uses ${HA_SBIN_DIR}/crm_master, so it exits
+#    127 without this override.  Fix: export HA_SBIN_DIR=/run/current-system/sw/bin.
+#
+# 3. Broad PATH for OCF agents: the resource executor (pacemaker-execd) runs
+#    OCF agent scripts as children.  NixOS provides no implicit PATH for
+#    system services; without an explicit PATH the agents can't find ip, ss,
+#    mount, umount, drbdadm, etc.
+#
+# 4. FUSER=true: the Filesystem OCF agent calls check_binary $FUSER (default:
+#    fuser from psmisc), which is not installed.  Setting FUSER=true makes
+#    check_binary succeed (true is always in PATH) and the subsequent
+#    "$FUSER -km $mountpoint" becomes a no-op.  Pair with force_unmount=false
+#    on each Filesystem resource unless you want lazy unmount behaviour.
+{ lib, pkgs, ... }:
+
+let
+  ocfBinPath = lib.concatStringsSep ":" [
+    "${pkgs.iproute2}/bin"
+    "${pkgs.iproute2}/sbin"
+    "${pkgs.iputils}/bin"
+    "${pkgs.util-linux}/bin"
+    "${pkgs.util-linux}/sbin"
+    "${pkgs.gawk}/bin"
+    "${pkgs.gnugrep}/bin"
+    "${pkgs.gnused}/bin"
+    "${pkgs.coreutils}/bin"
+    "${pkgs.bash}/bin"
+    "${pkgs.procps}/bin"
+    "${pkgs.xfsprogs}/bin"
+    "${pkgs.drbd}/bin"
+    "${pkgs.python3}/bin"
+    "/run/current-system/sw/bin"
+    "/run/current-system/sw/sbin"
+    "/usr/local/sbin"
+    "/usr/local/bin"
+    "/usr/sbin"
+    "/usr/bin"
+    "/sbin"
+    "/bin"
+  ];
+
+  # Single pre-start script: schemas symlink + directory ownership.
+  # Runs before pacemakerd so pacemaker-based finds hacluster-owned dirs.
+  preStartCmd = "${pkgs.bash}/bin/bash -c '"
+    + "ln -sfn ${pkgs.pacemaker}/share/pacemaker /var/lib/pacemaker/schemas; "
+    + "for d in /var/lib/pacemaker /var/lib/pacemaker/cib /var/lib/pacemaker/cores "
+    + "/var/lib/pacemaker/pengine /var/lib/pacemaker/blackbox "
+    + "/var/lib/pacemaker/hostcache; do "
+    + "mkdir -p \"\\$d\" && chown hacluster:pacemaker \"\\$d\" && chmod 2770 \"\\$d\"; "
+    + "done'";
+
+  ocfEnv = {
+    PATH = lib.mkForce ocfBinPath;
+    OCF_ROOT = "${pkgs.ocf-resource-agents}/usr/lib/ocf";
+    HA_SBIN_DIR = "/run/current-system/sw/bin";
+    FUSER = "true";
+  };
+in
+{
+  users.groups.haclient = { };
+
+  services.corosync.enable = true;
+  services.pacemaker.enable = true;
+
+  systemd.services = {
+    pacemaker = {
+      serviceConfig = {
+        StateDirectory = lib.mkForce "";
+        ExecStartPre = lib.mkBefore [ preStartCmd ];
+      };
+      environment = ocfEnv;
+    };
+    pacemaker-execd.environment = ocfEnv;
+  };
+
+  environment.systemPackages = with pkgs; [
+    corosync
+    pacemaker
+    ocf-resource-agents
+  ];
+}

modules/platforms/lxc.nix

@@ -63,23 +63,22 @@ in
     # back to decide `pct create`'s --unprivileged flag, so the two stay
     # in sync).
     #
-    # lxc-docker is the one exception: the kernel's NFS client doesn't set
+    # Any lxc-* host with an NFS fileSystem must be privileged: the kernel's
-    # FS_USERNS_MOUNT, so mounting NFS from inside *any* non-init user
+    # NFS client doesn't set FS_USERNS_MOUNT, so mounting NFS from inside
-    # namespace -- which is exactly what an unprivileged container's
+    # *any* non-init user namespace -- which is exactly what an unprivileged
-    # UID-mapped root runs in -- is rejected at the VFS layer with EPERM,
+    # container's UID-mapped root runs in -- is rejected at the VFS layer
-    # no matter what Proxmox's own `mount=nfs;nfs4` container feature
+    # with EPERM, no matter what Proxmox's own `mount=nfs;nfs4` container
-    # allows at the AppArmor layer (confirmed live: TCP to the NFS server
+    # feature allows at the AppArmor layer (confirmed live: TCP to the NFS
-    # succeeds, the server's export table matches the container's IP, and
+    # server succeeds, the server's export table matches the container's IP,
-    # `mount.nfs: Operation not permitted` still fires immediately with no
+    # and `mount.nfs: Operation not permitted` still fires immediately with
-    # corresponding denial anywhere in the server's logs -- a kernel-level
+    # no corresponding denial anywhere in the server's logs -- a kernel-level
-    # rejection, not a network or export-permission one). Keying off
+    # rejection, not a network or export-permission one). Deriving this from
-    # hostName rather than something docker-build-type-specific because
+    # fileSystems rather than a per-host override keeps it self-consistent:
-    # modules/build-types/docker.nix is also composed for linode-docker/
+    # any new lxc-* host that declares an NFS mount automatically gets the
-    # proxmox-docker, which don't import proxmox-lxc.nix at all --setting
+    # privilege level it needs without a separate manual flag.
-    # this option there would break their eval with "option does not
+    privileged = builtins.any
-    # exist" regardless of any mkIf guard, since mkIf only makes a value
+      (fs: fs.fsType == "nfs" || fs.fsType == "nfs4")
-    # conditional, not whether the option needs to exist somewhere.
+      (builtins.attrValues config.fileSystems);
-    privileged = config.networking.hostName == "docker";
   };
 
   boot.loader = {

modules/pxe-boot/mount-pxe-images.nix

@@ -0,0 +1,42 @@
+{ config, lib, vars, ... }:
+
+let
+  # Use the same FQDN approach as docker/mount-data.nix — a bare hostname is
+  # unreliable: systemd-resolved only tries LLMNR for single-label names, and
+  # a global search domain causes it to skip the interface-scoped LAN DNS.
+  nfsServer = "${vars.nfsServerHost}.${vars.homeDomain}";
+in
+{
+  fileSystems.${vars.nfsShares.pxebootImages.mountpoint} = {
+    device = "${nfsServer}:${vars.storageRoot}/${vars.nfsShares.pxebootImages.subpath}";
+    fsType = "nfs";
+    options = [
+      "_netdev"
+      "noatime"
+    ] ++ (if config.boot.isContainer
+    # NFSv4 requires rpc_pipefs (sunrpc filesystem), which Proxmox LXC
+    # containers block unless `features: mount=nfs` is set. Use NFSv3+nolock
+    # instead: no rpc_pipefs dependency at the protocol level, and rpcbind
+    # on the server handles port resolution without needing client-side
+    # sunrpc infrastructure. nofail keeps boot clean if server is unreachable.
+    then [ "nfsvers=3" "proto=tcp" "nolock" "nofail" ]
+    else [ "nfsvers=4.2" "x-systemd.automount" ]);
+  };
+
+  # NixOS pulls var-lib-nfs-rpc_pipefs.mount (the sunrpc filesystem) into
+  # nfs-client.target for any nfs fileSystems entry. In LXC containers the
+  # sunrpc mount is blocked by Proxmox's AppArmor profile, causing it to fail
+  # and the activation to report an error even though our mount uses nofail.
+  # Add ConditionVirtualization=!container via drop-in so systemd skips the
+  # unit entirely in containers (skip = inactive, not failed), which keeps
+  # nfs-client.target green and activation clean.
+  systemd.units = lib.mkIf config.boot.isContainer {
+    "var-lib-nfs-rpc_pipefs.mount" = {
+      overrideStrategy = "asDropin";
+      text = ''
+        [Unit]
+        ConditionVirtualization=!container
+      '';
+    };
+  };
+}

modules/tailscale/subnet-router.nix

@@ -1,4 +1,4 @@
-_:
+{ pkgs, ... }:
 
 {
   imports = [ ./enable-service.nix ];
@@ -12,4 +12,24 @@ _:
     # instead of relaying through DERP.
     openFirewall = true;
   };
+
+  # Tailscale recommends these ethtool flags on the uplink interface to get
+  # full UDP GRO throughput on subnet routers (https://tailscale.com/s/ethtool-config-udp-gro).
+  # The interface is derived from the default route so it works regardless of
+  # what the NIC is named on a given host.
+  systemd.services.tailscale-udp-gro = {
+    description = "Enable UDP GRO forwarding on uplink for Tailscale subnet router";
+    after = [ "network-online.target" ];
+    wants = [ "network-online.target" ];
+    wantedBy = [ "multi-user.target" ];
+    path = [ pkgs.ethtool pkgs.iproute2 ];
+    serviceConfig = {
+      Type = "oneshot";
+      RemainAfterExit = true;
+      ExecStart = pkgs.writeShellScript "tailscale-udp-gro" ''
+        NETDEV=$(ip -o route get 8.8.8.8 | cut -f 5 -d " ")
+        ethtool -K "$NETDEV" rx-udp-gro-forwarding on rx-gro-list off
+      '';
+    };
+  };
 }

modules/tailscale/ts-dns-forwarder.nix

@@ -0,0 +1,52 @@
+{ vars, ... }:
+
+{
+  # Run dnsmasq on the LAN interface as a forwarding-only resolver for
+  # *.ts.net (Tailscale MagicDNS names). FreeIPA's bind-dyndb-ldap
+  # cannot reach 100.100.100.100 (Tailscale's internal resolver) directly
+  # because the DC is not a Tailscale node. This host IS a Tailscale node
+  # and can reach 100.100.100.100 via its tailscale0 interface, so it
+  # acts as an intermediary: FreeIPA has a conditional forward zone for
+  # ts.net pointing here (vars.tailscaleRouterIp), and this dnsmasq
+  # instance forwards those queries onward to Tailscale's resolver.
+  #
+  # Configure FreeIPA once after deploying this host:
+  #   kinit admin
+  #   ipa dnsforwardzone-add ${vars.tailnetDomain} \
+  #     --forwarder=${vars.tailscaleRouterIp} \
+  #     --forward-policy=only
+  # Note: IPA refuses to shadow ts.net (a real public TLD); use the
+  # tailnet-specific subdomain (vars.tailnetDomain) instead.
+  services.dnsmasq = {
+    enable = true;
+    settings = {
+      # Listen only on the LAN interface — not tailscale0 or loopback.
+      # bind-interfaces prevents dnsmasq from binding to 0.0.0.0 and
+      # then filtering by interface later; combined with `interface` this
+      # ensures it genuinely listens only on eth0.
+      bind-interfaces = true;
+      interface = [ vars.lxcLanInterface ];
+
+      # Forward-only: no local /etc/hosts or /etc/resolv.conf reading,
+      # no negative caching of NXDOMAIN for names this instance doesn't
+      # serve. All ts.net queries come from FreeIPA's conditional forwarder
+      # and must be answered by Tailscale's resolver.
+      no-hosts = true;
+      no-resolv = true;
+
+      # Tailscale's internal "Quad100" resolver — reachable from any
+      # Tailscale node via the tailscale0 interface. Scoped to the
+      # specific tailnet subdomain (vars.tailnetDomain) rather than
+      # all of ts.net: FreeIPA refuses to shadow ts.net (a real public
+      # TLD with DNSimple nameservers) so the conditional forward zone
+      # in FreeIPA must use the tailnet-specific subdomain instead:
+      #   ipa dnsforwardzone-add ${vars.tailnetDomain} \
+      #     --forwarder=${vars.tailscaleRouterIp} \
+      #     --forward-policy=only
+      server = [ "/${vars.tailnetDomain}/100.100.100.100" ];
+    };
+  };
+
+  networking.firewall.allowedUDPPorts = [ 53 ];
+  networking.firewall.allowedTCPPorts = [ 53 ];
+}

scripts/ha/acceptance-tests.sh

@@ -0,0 +1,167 @@
+#!/usr/bin/env bash
+# acceptance-tests.sh — HA cluster acceptance tests (T1–T7)
+#
+# Run from a host with SSH access to both HA nodes (or from node1 itself).
+# All 7 tests must pass before considering the cluster production-ready.
+# Test values below must match variables.nix haServer* values.
+set -euo pipefail
+
+# ── Configuration ─────────────────────────────────────────────────────────
+NODE1="ha-server-1"
+NODE2="ha-server-2"
+NODE1_IP="192.168.2.200"   # vars.haServer1Ip
+NODE2_IP="192.168.2.201"   # vars.haServer2Ip
+VIP="192.168.2.202"        # vars.haServerVip
+XFS_MOUNT="/srv/ha-data"   # vars.haStorageRoot
+ISCSI_IQN="iqn.2026-01.home.sweet:ha-storage"  # vars.haIscsiIqn
+# ──────────────────────────────────────────────────────────────────────────
+
+PASS=0
+FAIL=0
+RESULTS=()
+
+pass() { echo "  PASS: $1"; ((PASS++)); RESULTS+=("PASS $1"); }
+fail() { echo "  FAIL: $1"; ((FAIL++)); RESULTS+=("FAIL $1"); }
+
+n1() { ssh -o StrictHostKeyChecking=no -o ConnectTimeout=5 "root@${NODE1_IP}" "$@" 2>/dev/null; }
+n2() { ssh -o StrictHostKeyChecking=no -o ConnectTimeout=5 "root@${NODE2_IP}" "$@" 2>/dev/null; }
+
+echo "════════════════════════════════════════════════════"
+echo " HA Cluster Acceptance Tests — $(date '+%Y-%m-%d %H:%M:%S')"
+echo "════════════════════════════════════════════════════"
+
+# ── T1: Corosync quorum established ──────────────────────────────────────
+echo ""
+echo "[T1] Corosync quorum"
+if n1 "corosync-quorumtool -s" 2>/dev/null | grep -q "Quorate:.*Yes"; then
+  pass "cluster has quorum"
+else
+  fail "cluster does not have quorum — check corosync on both nodes"
+fi
+
+# ── T2: DRBD Primary on node1, Secondary on node2 ────────────────────────
+echo ""
+echo "[T2] DRBD roles"
+DRBD_ROLE=$(n1 "drbdadm role ha-data" 2>/dev/null || echo "unknown")
+if [[ "$DRBD_ROLE" == "Primary/Secondary" || "$DRBD_ROLE" == "Primary" ]]; then
+  pass "DRBD Primary on $NODE1 ($DRBD_ROLE)"
+else
+  fail "unexpected DRBD role on $NODE1: $DRBD_ROLE (expected Primary/Secondary)"
+fi
+
+DRBD_DSTATE=$(n1 "drbdadm dstate ha-data" 2>/dev/null || echo "unknown")
+if echo "$DRBD_DSTATE" | grep -q "UpToDate"; then
+  pass "DRBD disk state UpToDate ($DRBD_DSTATE)"
+else
+  fail "DRBD disk not UpToDate: $DRBD_DSTATE"
+fi
+
+# ── T3: XFS mounted at haStorageRoot on the Active node ──────────────────
+echo ""
+echo "[T3] XFS mount"
+if n1 "mountpoint -q '${XFS_MOUNT}'" 2>/dev/null; then
+  pass "XFS mounted at ${XFS_MOUNT} on $NODE1"
+else
+  fail "XFS not mounted at ${XFS_MOUNT} on $NODE1"
+fi
+
+if n2 "mountpoint -q '${XFS_MOUNT}'" 2>/dev/null; then
+  fail "XFS unexpectedly mounted on $NODE2 (should only be on Active node)"
+else
+  pass "XFS not mounted on $NODE2 (correct — Secondary)"
+fi
+
+# ── T4: iSCSI target visible on both nodes ────────────────────────────────
+echo ""
+echo "[T4] iSCSI target"
+IQN_COUNT=$(n1 "ls /sys/kernel/config/target/iscsi/ 2>/dev/null | grep -c iqn" || echo "0")
+if [[ "$IQN_COUNT" -ge 1 ]]; then
+  pass "iSCSI IQN active on $NODE1 ($IQN_COUNT target(s))"
+else
+  fail "no iSCSI IQN active on $NODE1"
+fi
+
+# iSCSI discovery from node2 via VIP
+if n2 "iscsiadm -m discovery -t sendtargets -p '${VIP}' 2>/dev/null | grep -q '${ISCSI_IQN}'"; then
+  pass "iSCSI target discoverable from $NODE2 via VIP ${VIP}"
+else
+  fail "iSCSI target not discoverable from $NODE2 via ${VIP}"
+fi
+
+# ── T5: Failover — standby node1, verify resources move to node2 ──────────
+echo ""
+echo "[T5] Failover (standby $NODE1)"
+MYNODE=$(n1 "crm_node -n" 2>/dev/null || echo "")
+n1 "crm_standby -N '${MYNODE}' -v on" 2>/dev/null || true
+echo "  Waiting up to 30 s for resources to move to $NODE2..."
+MOVED=false
+for i in $(seq 1 30); do
+  if n2 "mountpoint -q '${XFS_MOUNT}'" 2>/dev/null; then
+    MOVED=true
+    echo "  Resources moved in ${i}s"
+    break
+  fi
+  sleep 1
+done
+
+if $MOVED; then
+  pass "XFS mounted on $NODE2 after failover"
+  IQN_ON_N2=$(n2 "ls /sys/kernel/config/target/iscsi/ 2>/dev/null | grep -c iqn" || echo "0")
+  [[ "$IQN_ON_N2" -ge 1 ]] \
+    && pass "iSCSI target active on $NODE2 after failover" \
+    || fail "iSCSI target NOT active on $NODE2 after failover"
+else
+  fail "XFS did not mount on $NODE2 within 30 s — failover incomplete"
+fi
+
+# ── T6: Data integrity — file written pre-failover readable post-failover ─
+echo ""
+echo "[T6] Data integrity"
+# Write a test file on node2 (now Active) and verify its content
+TEST_FILE="${XFS_MOUNT}/.acceptance-test-$$"
+TEST_CONTENT="ha-acceptance-test-$(date +%s)"
+n2 "echo '${TEST_CONTENT}' > '${TEST_FILE}'" 2>/dev/null || true
+READBACK=$(n2 "cat '${TEST_FILE}' 2>/dev/null" || echo "")
+if [[ "$READBACK" == "$TEST_CONTENT" ]]; then
+  pass "test file written and read back correctly on $NODE2"
+else
+  fail "data integrity check failed (wrote: '$TEST_CONTENT', read: '$READBACK')"
+fi
+n2 "rm -f '${TEST_FILE}'" 2>/dev/null || true
+
+# ── T7: Node rejoin — un-standby node1, verify cluster is healthy ─────────
+echo ""
+echo "[T7] Node rejoin"
+n1 "crm_standby -N '${MYNODE}' -v off" 2>/dev/null || true
+n1 "crm_resource --cleanup" 2>/dev/null || true
+sleep 5
+
+ONLINE_NODES=$(n2 "crm_mon -1 2>/dev/null | grep -c 'Online:'" || echo "0")
+if n1 "corosync-quorumtool -s 2>/dev/null | grep -q 'Quorate:.*Yes'"; then
+  pass "$NODE1 rejoined — cluster has quorum"
+else
+  fail "$NODE1 did not rejoin with quorum"
+fi
+
+DRBD_ROLE_AFTER=$(n1 "drbdadm role ha-data" 2>/dev/null || echo "unknown")
+if echo "$DRBD_ROLE_AFTER" | grep -q "Secondary"; then
+  pass "$NODE1 is DRBD Secondary after rejoin ($DRBD_ROLE_AFTER)"
+else
+  fail "unexpected DRBD role on $NODE1 after rejoin: $DRBD_ROLE_AFTER"
+fi
+
+# ── Summary ───────────────────────────────────────────────────────────────
+echo ""
+echo "════════════════════════════════════════════════════"
+echo " Results: ${PASS} PASS, ${FAIL} FAIL"
+echo "════════════════════════════════════════════════════"
+for r in "${RESULTS[@]}"; do echo "  $r"; done
+echo ""
+
+if [[ "$FAIL" -eq 0 ]]; then
+  echo "ALL PASS — cluster is production-ready."
+  exit 0
+else
+  echo "SOME TESTS FAILED — investigate before deploying."
+  exit 1
+fi

scripts/ha/cluster-enable-stonith.sh

@@ -0,0 +1,86 @@
+#!/usr/bin/env bash
+# cluster-enable-stonith.sh — enable STONITH fence agent after the fence SSH
+# key is deployed to both nodes and authorised on the Proxmox host.
+#
+# Run from ha-server-1 as root AFTER:
+#   - /etc/pacemaker/fence_pve_ssh exists on both nodes (chmod +x)
+#     (copy from scripts/ha/fence-pve-ssh.py)
+#   - /etc/fence-pve-ssh-key (SSH private key) exists on both nodes
+#   - The corresponding public key is in authorized_keys on PVE_HOST
+#   - VMID_NODE1 / VMID_NODE2 filled in below
+set -euo pipefail
+
+# ── Configuration ─────────────────────────────────────────────────────────
+NODE1="ha-server-1"
+NODE2="ha-server-2"
+VMID_NODE1=""          # FILL IN: Proxmox VMID for ha-server-1
+VMID_NODE2=""          # FILL IN: Proxmox VMID for ha-server-2
+PVE_HOST="pve1.sweet.home"
+PVE_USER="wayne"
+FENCE_KEY="/etc/fence-pve-ssh-key"
+FENCE_SCRIPT="/etc/pacemaker/fence_pve_ssh"
+# ──────────────────────────────────────────────────────────────────────────
+
+log() { echo "[stonith-setup] $*"; }
+die() { echo "[stonith-setup] ERROR: $*" >&2; exit 1; }
+
+[[ $(id -u) -eq 0 ]] || die "must run as root"
+[[ -n "$VMID_NODE1" ]] || die "VMID_NODE1 not set — edit this script"
+[[ -n "$VMID_NODE2" ]] || die "VMID_NODE2 not set — edit this script"
+[[ -f "$FENCE_KEY" ]] || die "fence key not found at $FENCE_KEY"
+[[ -f "$FENCE_SCRIPT" ]] || die "fence script not found at $FENCE_SCRIPT"
+
+log "Verifying fence agent can reach ${PVE_HOST}..."
+ssh -i "$FENCE_KEY" -o BatchMode=yes -o ConnectTimeout=10 \
+    -o StrictHostKeyChecking=no "${PVE_USER}@${PVE_HOST}" \
+    "sudo /usr/sbin/qm list" &>/dev/null \
+  || die "Cannot SSH to ${PVE_USER}@${PVE_HOST} — check authorized_keys and sudo"
+log "Fence agent SSH connectivity confirmed"
+
+log "Creating Pacemaker STONITH resources..."
+cibadmin --create --scope resources --xml-text "
+<primitive id=\"stonith-${NODE1}\" class=\"stonith\" type=\"external/fence_pve_ssh\">
+  <instance_attributes id=\"stonith-${NODE1}-attrs\">
+    <nvpair id=\"stonith-${NODE1}-plug\" name=\"plug\" value=\"${NODE1}\"/>
+    <nvpair id=\"stonith-${NODE1}-pve-host\" name=\"pve_host\" value=\"${PVE_HOST}\"/>
+    <nvpair id=\"stonith-${NODE1}-pve-user\" name=\"pve_user\" value=\"${PVE_USER}\"/>
+    <nvpair id=\"stonith-${NODE1}-key-file\" name=\"key_file\" value=\"${FENCE_KEY}\"/>
+    <nvpair id=\"stonith-${NODE1}-vmid1\" name=\"vmid_node1\" value=\"${VMID_NODE1}\"/>
+    <nvpair id=\"stonith-${NODE1}-vmid2\" name=\"vmid_node2\" value=\"${VMID_NODE2}\"/>
+    <nvpair id=\"stonith-${NODE1}-host-list\" name=\"pcmk_host_list\" value=\"${NODE1}\"/>
+  </instance_attributes>
+  <operations>
+    <op id=\"stonith-${NODE1}-monitor\" name=\"monitor\" interval=\"30s\" timeout=\"30s\"/>
+  </operations>
+</primitive>
+" 2>/dev/null || true
+
+cibadmin --create --scope resources --xml-text "
+<primitive id=\"stonith-${NODE2}\" class=\"stonith\" type=\"external/fence_pve_ssh\">
+  <instance_attributes id=\"stonith-${NODE2}-attrs\">
+    <nvpair id=\"stonith-${NODE2}-plug\" name=\"plug\" value=\"${NODE2}\"/>
+    <nvpair id=\"stonith-${NODE2}-pve-host\" name=\"pve_host\" value=\"${PVE_HOST}\"/>
+    <nvpair id=\"stonith-${NODE2}-pve-user\" name=\"pve_user\" value=\"${PVE_USER}\"/>
+    <nvpair id=\"stonith-${NODE2}-key-file\" name=\"key_file\" value=\"${FENCE_KEY}\"/>
+    <nvpair id=\"stonith-${NODE2}-vmid1\" name=\"vmid_node1\" value=\"${VMID_NODE1}\"/>
+    <nvpair id=\"stonith-${NODE2}-vmid2\" name=\"vmid_node2\" value=\"${VMID_NODE2}\"/>
+    <nvpair id=\"stonith-${NODE2}-host-list\" name=\"pcmk_host_list\" value=\"${NODE2}\"/>
+  </instance_attributes>
+  <operations>
+    <op id=\"stonith-${NODE2}-monitor\" name=\"monitor\" interval=\"30s\" timeout=\"30s\"/>
+  </operations>
+</primitive>
+" 2>/dev/null || true
+
+log "Enabling STONITH and restoring quorum policy..."
+crm_attribute -t crm_config -n stonith-enabled -v true
+crm_attribute -t crm_config -n no-quorum-policy -v stop
+
+log "DRBD fencing mode must also be updated to resource-only (already the"
+log "default in cluster-config.nix; confirm with: cat /etc/drbd.d/ha-data.conf)"
+
+log "Testing fence agent..."
+stonith_admin --list-devices && log "Fence devices listed successfully." \
+  || warn "stonith_admin --list-devices failed — check config"
+
+log "STONITH enabled. Cluster is now fully HA."

scripts/ha/cluster-init.sh

@@ -0,0 +1,284 @@
+#!/usr/bin/env bash
+# cluster-init.sh — one-time HA cluster initialisation script
+#
+# Run ONCE from ha-server-1 as root AFTER both VMs are booted and have SSH
+# access.  It:
+#   1. Generates and distributes the corosync authkey
+#   2. Waits for corosync quorum and pacemaker
+#   3. Initialises DRBD metadata, promotes node1 to primary
+#   4. Creates XFS on /dev/drbd0 and mounts it
+#   5. Creates the directory tree and iSCSI LUN backing file
+#   6. Configures LIO iSCSI target (file-backed LUN)
+#   7. Configures Pacemaker resources: DRBD → XFS → iSCSI → NFS → VIP
+#
+# Prerequisites:
+#   - Both VMs booted with the ha-server config (nixos-rebuild done)
+#   - SSH key access from node1 to root@NODE2_IP
+#   - VMID_NODE1 / VMID_NODE2 filled in below (needed for STONITH setup;
+#     cluster starts without STONITH, which you enable separately via
+#     scripts/ha/cluster-enable-stonith.sh)
+#   - Run as root on ha-server-1
+set -euo pipefail
+
+# ── Configuration ─────────────────────────────────────────────────────────
+# These must match variables.nix haServer* values and the Proxmox VMID
+# assignments.  Update before running.
+NODE1="ha-server-1"
+NODE2="ha-server-2"
+NODE1_IP="192.168.2.200"   # vars.haServer1Ip
+NODE2_IP="192.168.2.201"   # vars.haServer2Ip
+VIP="192.168.2.202"        # vars.haServerVip
+XFS_MOUNT="/srv/ha-data"   # vars.haStorageRoot
+ISCSI_IQN="iqn.2026-01.home.sweet:ha-storage"  # vars.haIscsiIqn
+ISCSI_LUN_FILE="${XFS_MOUNT}/iscsi-lun.img"
+ISCSI_LUN_SIZE="10G"
+DRBD_DEVICE="/dev/drbd0"
+VMID_NODE1=""              # FILL IN: Proxmox VMID for ha-server-1
+VMID_NODE2=""              # FILL IN: Proxmox VMID for ha-server-2
+PVE_HOST="pve1.sweet.home"
+PVE_USER="wayne"
+
+# NFS dataset subdirectories to create under XFS_MOUNT.
+# Must mirror vars.nfsShares subpath values in variables.nix.
+NFS_SUBDIRS=(
+  "docker/config"
+  "docker/volumes"
+  "docker/databases"
+  "docker/nextcloud-data"
+  "raspi/volumes"
+  "proxmox/iso"
+  "proxmox/lxc"
+  "pxe-boot/images"
+)
+# ──────────────────────────────────────────────────────────────────────────
+
+log()  { echo "[cluster-init] $*"; }
+die()  { echo "[cluster-init] ERROR: $*" >&2; exit 1; }
+warn() { echo "[cluster-init] WARNING: $*" >&2; }
+
+[[ $(id -u) -eq 0 ]] || die "must run as root"
+[[ "$(hostname)" == "$NODE1" ]] || die "must run on $NODE1"
+
+# ── 0. Corosync authkey ───────────────────────────────────────────────────
+AUTHKEY="/etc/corosync/authkey"
+mkdir -p /etc/corosync
+if [[ ! -f "$AUTHKEY" ]]; then
+  log "Generating corosync authkey..."
+  corosync-keygen -k "$AUTHKEY"
+  chmod 0400 "$AUTHKEY"
+fi
+log "Distributing authkey to $NODE2..."
+ssh "root@${NODE2_IP}" "mkdir -p /etc/corosync"
+scp -q "$AUTHKEY" "root@${NODE2_IP}:${AUTHKEY}"
+ssh "root@${NODE2_IP}" "chmod 0400 '${AUTHKEY}'"
+
+log "Restarting corosync on both nodes..."
+systemctl restart corosync
+ssh "root@${NODE2_IP}" "systemctl restart corosync"
+sleep 3
+
+# ── 1. Corosync quorum ────────────────────────────────────────────────────
+log "Waiting for corosync quorum..."
+for i in $(seq 1 30); do
+  if corosync-quorumtool -s 2>/dev/null | grep -q 'Quorate:.*Yes'; then
+    log "Quorum established"
+    break
+  fi
+  [[ $i -eq 30 ]] && die "corosync quorum not established after 60 s"
+  sleep 2
+done
+
+log "Waiting for pacemaker..."
+for i in $(seq 1 30); do
+  if crm_mon -1 &>/dev/null; then
+    log "Pacemaker running"
+    break
+  fi
+  [[ $i -eq 30 ]] && die "pacemaker not running after 60 s"
+  sleep 2
+done
+
+# ── 2. DRBD initialisation ────────────────────────────────────────────────
+log "Initialising DRBD metadata on $NODE1..."
+if ! drbdadm dstate ha-data 2>/dev/null | grep -q "UpToDate\|Inconsistent\|Diskless"; then
+  drbdadm create-md ha-data --force
+fi
+
+log "Initialising DRBD metadata on $NODE2..."
+ssh "root@${NODE2_IP}" "
+  if ! drbdadm dstate ha-data 2>/dev/null | grep -q 'UpToDate\|Inconsistent\|Diskless'; then
+    drbdadm create-md ha-data --force
+  fi
+"
+
+log "Bringing up DRBD on both nodes..."
+drbdadm up ha-data 2>/dev/null || true
+ssh "root@${NODE2_IP}" "drbdadm up ha-data 2>/dev/null" || true
+
+log "Forcing $NODE1 to DRBD Primary for initial sync..."
+drbdadm primary ha-data --force
+
+log "Waiting for DRBD to finish initial sync (this may take several minutes)..."
+for i in $(seq 1 300); do
+  state=$(drbdadm dstate ha-data 2>/dev/null || echo "unknown")
+  if echo "$state" | grep -q "UpToDate/UpToDate"; then
+    log "DRBD sync complete: $state"
+    break
+  fi
+  [[ $i -eq 300 ]] && warn "DRBD not UpToDate after 300 s — continuing anyway (check drbdadm status)"
+  sleep 1
+done
+
+# ── 3. XFS filesystem ─────────────────────────────────────────────────────
+log "Creating XFS on ${DRBD_DEVICE}..."
+if ! xfs_info "${DRBD_DEVICE}" &>/dev/null; then
+  mkfs.xfs -f "${DRBD_DEVICE}"
+fi
+
+log "Mounting ${DRBD_DEVICE} at ${XFS_MOUNT}..."
+mkdir -p "${XFS_MOUNT}"
+mount "${DRBD_DEVICE}" "${XFS_MOUNT}"
+
+# ── 4. NFS dataset directories ────────────────────────────────────────────
+log "Creating NFS dataset directories..."
+for subdir in "${NFS_SUBDIRS[@]}"; do
+  mkdir -p "${XFS_MOUNT}/${subdir}"
+done
+
+# ── 5. iSCSI LUN backing file ─────────────────────────────────────────────
+log "Creating iSCSI LUN backing file ${ISCSI_LUN_FILE} (${ISCSI_LUN_SIZE})..."
+if [[ ! -f "${ISCSI_LUN_FILE}" ]]; then
+  fallocate -l "${ISCSI_LUN_SIZE}" "${ISCSI_LUN_FILE}"
+fi
+
+# ── 6. LIO iSCSI target ───────────────────────────────────────────────────
+log "Configuring LIO iSCSI target via targetcli..."
+targetcli <<EOF
+/backstores/fileio create name=ha-lun0 file_or_dev=${ISCSI_LUN_FILE} size=0 write_back=false
+/iscsi create ${ISCSI_IQN}
+/iscsi/${ISCSI_IQN}/tpg1/luns create /backstores/fileio/ha-lun0
+/iscsi/${ISCSI_IQN}/tpg1/portals create ${VIP}
+/iscsi/${ISCSI_IQN}/tpg1 set attribute authentication=0
+/iscsi/${ISCSI_IQN}/tpg1 set attribute demo_mode_write_protect=0
+saveconfig /etc/target/saveconfig.json
+EOF
+
+log "Distributing iSCSI saveconfig to $NODE2..."
+scp -q /etc/target/saveconfig.json "root@${NODE2_IP}:/etc/target/saveconfig.json"
+
+log "Unmounting ${XFS_MOUNT} — Pacemaker manages it..."
+umount "${XFS_MOUNT}"
+
+log "Demoting DRBD to Secondary — Pacemaker manages primary role..."
+drbdadm secondary ha-data
+
+# ── 7. Pacemaker resources ────────────────────────────────────────────────
+log "Configuring Pacemaker cluster properties..."
+crm_attribute -t crm_config -n stonith-enabled -v false
+crm_attribute -t crm_config -n no-quorum-policy -v ignore
+
+log "Creating DRBD promotable clone resource..."
+cibadmin --replace --scope resources --xml-text "
+<resources>
+  <clone id=\"ms-drbd0\" globally-unique=\"false\">
+    <meta_attributes id=\"ms-drbd0-meta\">
+      <nvpair id=\"ms-drbd0-promotable\" name=\"promotable\" value=\"true\"/>
+      <nvpair id=\"ms-drbd0-master-max\" name=\"master-max\" value=\"1\"/>
+      <nvpair id=\"ms-drbd0-master-node-max\" name=\"master-node-max\" value=\"1\"/>
+      <nvpair id=\"ms-drbd0-clone-max\" name=\"clone-max\" value=\"2\"/>
+      <nvpair id=\"ms-drbd0-clone-node-max\" name=\"clone-node-max\" value=\"1\"/>
+      <nvpair id=\"ms-drbd0-notify\" name=\"notify\" value=\"true\"/>
+      <nvpair id=\"ms-drbd0-interleave\" name=\"interleave\" value=\"true\"/>
+    </meta_attributes>
+    <primitive id=\"drbd0\" class=\"ocf\" type=\"drbd\" provider=\"linbit\">
+      <instance_attributes id=\"drbd0-attrs\">
+        <nvpair id=\"drbd0-resource\" name=\"drbd_resource\" value=\"ha-data\"/>
+      </instance_attributes>
+      <operations>
+        <op id=\"drbd0-start\" name=\"start\" interval=\"0\" timeout=\"240s\"/>
+        <op id=\"drbd0-stop\" name=\"stop\" interval=\"0\" timeout=\"120s\"/>
+        <op id=\"drbd0-promote\" name=\"promote\" interval=\"0\" timeout=\"90s\"/>
+        <op id=\"drbd0-demote\" name=\"demote\" interval=\"0\" timeout=\"90s\"/>
+        <op id=\"drbd0-monitor-master\" name=\"monitor\" interval=\"20s\" timeout=\"20s\" role=\"Promoted\"/>
+        <op id=\"drbd0-monitor-slave\" name=\"monitor\" interval=\"30s\" timeout=\"20s\" role=\"Unpromoted\"/>
+      </operations>
+    </primitive>
+  </clone>
+  <group id=\"ha-group\">
+    <primitive id=\"xfs-data\" class=\"ocf\" type=\"Filesystem\" provider=\"heartbeat\">
+      <instance_attributes id=\"xfs-data-attrs\">
+        <nvpair id=\"xfs-data-device\" name=\"device\" value=\"${DRBD_DEVICE}\"/>
+        <nvpair id=\"xfs-data-directory\" name=\"directory\" value=\"${XFS_MOUNT}\"/>
+        <nvpair id=\"xfs-data-fstype\" name=\"fstype\" value=\"xfs\"/>
+        <nvpair id=\"xfs-data-options\" name=\"options\" value=\"defaults\"/>
+        <nvpair id=\"xfs-data-force_unmount\" name=\"force_unmount\" value=\"false\"/>
+      </instance_attributes>
+      <operations>
+        <op id=\"xfs-data-start\" name=\"start\" interval=\"0\" timeout=\"60s\"/>
+        <op id=\"xfs-data-stop\" name=\"stop\" interval=\"0\" timeout=\"60s\"/>
+        <op id=\"xfs-data-monitor\" name=\"monitor\" interval=\"20s\" timeout=\"40s\"/>
+      </operations>
+    </primitive>
+    <primitive id=\"iscsi-target\" class=\"systemd\" type=\"targetctl\">
+      <operations>
+        <op id=\"iscsi-start\" name=\"start\" interval=\"0\" timeout=\"60s\"/>
+        <op id=\"iscsi-stop\" name=\"stop\" interval=\"0\" timeout=\"60s\"/>
+        <op id=\"iscsi-monitor\" name=\"monitor\" interval=\"20s\" timeout=\"40s\"/>
+      </operations>
+    </primitive>
+    <primitive id=\"nfs-server\" class=\"systemd\" type=\"nfs-server\">
+      <operations>
+        <op id=\"nfs-start\" name=\"start\" interval=\"0\" timeout=\"60s\"/>
+        <op id=\"nfs-stop\" name=\"stop\" interval=\"0\" timeout=\"60s\"/>
+        <op id=\"nfs-monitor\" name=\"monitor\" interval=\"30s\" timeout=\"40s\"/>
+      </operations>
+    </primitive>
+    <primitive id=\"vip\" class=\"ocf\" type=\"IPaddr2\" provider=\"heartbeat\">
+      <instance_attributes id=\"vip-attrs\">
+        <nvpair id=\"vip-ip\" name=\"ip\" value=\"${VIP}\"/>
+        <nvpair id=\"vip-cidr\" name=\"cidr_netmask\" value=\"24\"/>
+      </instance_attributes>
+      <operations>
+        <op id=\"vip-start\" name=\"start\" interval=\"0\" timeout=\"20s\"/>
+        <op id=\"vip-stop\" name=\"stop\" interval=\"0\" timeout=\"20s\"/>
+        <op id=\"vip-monitor\" name=\"monitor\" interval=\"10s\" timeout=\"20s\"/>
+      </operations>
+    </primitive>
+  </group>
+</resources>
+"
+
+log "Adding ordering and colocation constraints..."
+cibadmin --create --scope constraints --xml-text "
+<constraints>
+  <rsc_order id=\"order-drbd-group\" first=\"ms-drbd0\" first-action=\"promote\" then=\"ha-group\" then-action=\"start\"/>
+  <rsc_colocation id=\"coloc-group-with-drbd\" rsc=\"ha-group\" with-rsc=\"ms-drbd0\" with-rsc-role=\"Master\" score=\"INFINITY\"/>
+</constraints>
+"
+
+log "Waiting for resources to start..."
+for i in $(seq 1 60); do
+  if crm_resource -r vip --locate 2>/dev/null | grep -q "running on"; then
+    log "VIP is up: $(crm_resource -r vip --locate)"
+    break
+  fi
+  [[ $i -eq 60 ]] && { warn "VIP not up after 120 s — check: crm_mon -1"; break; }
+  sleep 2
+done
+
+log ""
+log "═══════════════════════════════════════════════════════════════"
+log " HA cluster initialised."
+log ""
+log "  crm_mon -1                         — cluster status"
+log "  iscsiadm -m discovery -t st -p ${VIP}  — verify iSCSI target"
+log "  showmount -e ${VIP}                    — verify NFS exports"
+log ""
+log "  To enable STONITH (after deploying fence SSH key):"
+log "    1. Fill in VMID_NODE1 / VMID_NODE2 in cluster-enable-stonith.sh"
+log "    2. Copy scripts/ha/fence-pve-ssh.py to /etc/pacemaker/fence_pve_ssh"
+log "       on both nodes (chmod +x)"
+log "    3. Generate and distribute the fence SSH key"
+log "       (see docs or cluster-enable-stonith.sh header)"
+log "    4. bash scripts/ha/cluster-enable-stonith.sh"
+log "═══════════════════════════════════════════════════════════════"

scripts/ha/fence-pve-ssh.py

@@ -0,0 +1,179 @@
+#!/usr/bin/env python3
+"""
+fence_pve_ssh - Proxmox VE SSH fence agent for Pacemaker.
+
+Uses SSH to reach the Proxmox host and run 'qm stop/start <vmid>'.
+Deploy to /etc/pacemaker/fence_pve_ssh on both HA nodes (chmod +x).
+
+Configuration (as pacemaker stonith resource attributes):
+  pve_host     Proxmox host to SSH to (default: pve1.sweet.home)
+  pve_user     SSH user (default: wayne)
+  key_file     SSH private key path (default: /etc/fence-pve-ssh-key)
+  vmid_node1   VMID for ha-server-1
+  vmid_node2   VMID for ha-server-2
+  plug         Node name to act on (set by pacemaker: ha-server-1 or ha-server-2)
+  action       Action: off|on|reboot|status|list|metadata
+"""
+
+import argparse
+import subprocess
+import sys
+import os
+
+
+METADATA = """<?xml version="1.0" ?>
+<resource-agent name="fence_pve_ssh" shortdesc="Proxmox VE SSH fence agent (test lab)">
+  <longdesc>Fences a VM on a Proxmox VE host by SSHing to the PVE host and
+  running qm stop/start.  For test use only.</longdesc>
+  <vendor-url>https://proxmox.com</vendor-url>
+  <parameters>
+    <parameter name="action" required="1" unique="0">
+      <getopt mixed="-a, --action=[action]"/>
+      <content type="string" default="reboot"/>
+      <shortdesc lang="en">Fencing action: off|on|reboot|status|list</shortdesc>
+    </parameter>
+    <parameter name="plug" required="0" unique="0">
+      <getopt mixed="-n, --plug=[nodename]"/>
+      <content type="string"/>
+      <shortdesc lang="en">Cluster node name to fence</shortdesc>
+    </parameter>
+    <parameter name="pve_host" required="0" unique="0">
+      <getopt mixed="--pve-host=[host]"/>
+      <content type="string" default="pve1.sweet.home"/>
+      <shortdesc lang="en">Proxmox VE host to SSH to</shortdesc>
+    </parameter>
+    <parameter name="pve_user" required="0" unique="0">
+      <getopt mixed="--pve-user=[user]"/>
+      <content type="string" default="wayne"/>
+      <shortdesc lang="en">SSH user on the Proxmox host</shortdesc>
+    </parameter>
+    <parameter name="key_file" required="0" unique="0">
+      <getopt mixed="--key-file=[path]"/>
+      <content type="string" default="/etc/fence-pve-ssh-key"/>
+      <shortdesc lang="en">SSH private key file path</shortdesc>
+    </parameter>
+    <parameter name="vmid_node1" required="1" unique="0">
+      <getopt mixed="--vmid-node1=[vmid]"/>
+      <content type="string"/>
+      <shortdesc lang="en">VMID for ha-test-node1</shortdesc>
+    </parameter>
+    <parameter name="vmid_node2" required="1" unique="0">
+      <getopt mixed="--vmid-node2=[vmid]"/>
+      <content type="string"/>
+      <shortdesc lang="en">VMID for ha-test-node2</shortdesc>
+    </parameter>
+  </parameters>
+  <actions>
+    <action name="off" timeout="60s"/>
+    <action name="on" timeout="60s"/>
+    <action name="reboot" timeout="60s"/>
+    <action name="status" timeout="30s"/>
+    <action name="list" timeout="10s"/>
+    <action name="metadata" timeout="5s"/>
+  </actions>
+</resource-agent>
+"""
+
+
+def parse_args():
+    p = argparse.ArgumentParser(add_help=False)
+    p.add_argument("-a", "--action", default="reboot")
+    p.add_argument("-n", "--plug")
+    p.add_argument("--pve-host", default="pve1.sweet.home")
+    p.add_argument("--pve-user", default="wayne")
+    p.add_argument("--key-file", default="/etc/fence-pve-ssh-key")
+    p.add_argument("--vmid-node1")
+    p.add_argument("--vmid-node2")
+    # Allow remaining unknown args (pacemaker may pass extra ones)
+    return p.parse_known_args()[0]
+
+
+def ssh(pve_host, pve_user, key_file, cmd):
+    result = subprocess.run(
+        [
+            "ssh",
+            "-i", key_file,
+            "-o", "StrictHostKeyChecking=no",
+            "-o", "BatchMode=yes",
+            "-o", "ConnectTimeout=10",
+            f"{pve_user}@{pve_host}",
+            cmd,
+        ],
+        capture_output=True,
+        text=True,
+        timeout=30,
+    )
+    return result
+
+
+def get_vmid(args):
+    node = args.plug
+    if not node:
+        print("ERROR: --plug not specified", file=sys.stderr)
+        sys.exit(1)
+    mapping = {
+        "ha-server-1": args.vmid_node1,
+        "ha-server-2": args.vmid_node2,
+    }
+    vmid = mapping.get(node)
+    if not vmid:
+        print(f"ERROR: unknown node '{node}'", file=sys.stderr)
+        sys.exit(1)
+    return vmid
+
+
+def main():
+    args = parse_args()
+    action = args.action.lower()
+
+    if action == "metadata":
+        print(METADATA)
+        sys.exit(0)
+
+    if action == "list":
+        if args.vmid_node1:
+            print("ha-server-1")
+        if args.vmid_node2:
+            print("ha-server-2")
+        sys.exit(0)
+
+    vmid = get_vmid(args)
+
+    if not os.path.exists(args.key_file):
+        print(f"ERROR: SSH key not found at {args.key_file}", file=sys.stderr)
+        sys.exit(1)
+
+    if action in ("off", "reboot"):
+        print(f"Stopping VM {vmid} ({args.plug}) on {args.pve_host}...")
+        r = ssh(args.pve_host, args.pve_user, args.key_file,
+                f"sudo /usr/sbin/qm stop {vmid}")
+        if r.returncode != 0:
+            print(f"ERROR stopping VM: {r.stderr}", file=sys.stderr)
+            sys.exit(1)
+        print(f"VM {vmid} stopped")
+
+    if action in ("on", "reboot"):
+        print(f"Starting VM {vmid} ({args.plug}) on {args.pve_host}...")
+        r = ssh(args.pve_host, args.pve_user, args.key_file,
+                f"sudo /usr/sbin/qm start {vmid}")
+        if r.returncode != 0:
+            print(f"ERROR starting VM: {r.stderr}", file=sys.stderr)
+            sys.exit(1)
+        print(f"VM {vmid} started")
+
+    if action == "status":
+        r = ssh(args.pve_host, args.pve_user, args.key_file,
+                f"sudo /usr/sbin/qm status {vmid}")
+        if r.returncode != 0:
+            print(f"ERROR querying VM status: {r.stderr}", file=sys.stderr)
+            sys.exit(1)
+        # qm status returns "status: running" or "status: stopped"
+        status_line = r.stdout.strip()
+        print(status_line)
+        if "stopped" in status_line:
+            sys.exit(2)   # pacemaker interprets exit 2 as "off"
+        sys.exit(0)       # running = exit 0
+
+
+if __name__ == "__main__":
+    main()

scripts/proxmox/create-proxmox-resource.sh

@@ -867,17 +867,17 @@ if [[ "$type" == "lxc" ]]; then
   local_swap="${swap:-$memory}"
   # --unprivileged: read back from modules/platforms/lxc.nix's own
   # proxmoxLXC.privileged (via flake_target_lxc_privileged) rather than
-  # hardcoded, since that's no longer the same for every lxc-* target --
+  # hardcoded. lxc.nix derives this automatically: any lxc-* host whose
-  # lxc-docker sets it true so the container's NFS mounts work at all (the
+  # config.fileSystems has an NFS entry gets privileged=true, because the
-  # kernel's NFS client can't mount from inside any unprivileged
+  # kernel's NFS client (FS_USERNS_MOUNT not set) rejects NFS mounts from
-  # container's user namespace, no matter what AppArmor allows -- see that
+  # inside any non-init user namespace -- exactly what an unprivileged
-  # option's own comment). The NixOS config inside the image bakes in
+  # container's UID-mapped root lives in -- with EPERM at the VFS layer,
-  # cgroup/capability/mount expectations matching whichever value it was
+  # regardless of AppArmor (see lxc.nix's own comment). The NixOS config
-  # built with, so this must stay in sync with it -- `pct create`'s own
+  # bakes in cgroup/capability/mount expectations matching whichever value
-  # CLI default for this flag is privileged (unlike the web UI, which
+  # it was built with, so this must stay in sync -- `pct create`'s CLI
-  # defaults its checkbox the other way), so leaving it unset would create
+  # default is privileged (unlike the web UI, which defaults the other
-  # a privileged container running a NixOS config that assumes
+  # way), so leaving it unset would create a privileged container running
-  # unprivileged for every target except lxc-docker, a real mismatch.
+  # a NixOS config that assumes unprivileged, a real mismatch.
   privileged_eval="$(flake_target_lxc_privileged "$repo_root" "$flake_target")"
   unprivileged_flag=1
   [[ "$privileged_eval" == "true" ]] && unprivileged_flag=0

scripts/secrets/sync-host-keys.sh

@@ -85,6 +85,10 @@ ensure_admin_decrypt_key() {
   fi
 
   local key_file="$DEFAULT_SOPS_AGE_KEY_FILE"
+  # Expand a leading ~ that survived variable substitution without tilde
+  # expansion (happens when SOPS_AGE_KEY_FILE or XDG_CONFIG_HOME is set with
+  # a literal ~ in the caller's environment).
+  key_file="${key_file/#~\//$HOME/}"
 
   if [[ -s "$key_file" ]]; then
     echo "Found existing sops age key at ${key_file}."
@@ -93,36 +97,23 @@ ensure_admin_decrypt_key() {
 
   if [[ "$dry_run" -eq 1 ]]; then
     echo "[dry-run] No sops age decryption key found (checked \$SOPS_AGE_KEY, \$SOPS_AGE_KEY_FILE, ${key_file})."
-    echo "[dry-run] Would generate a new one here -- continuing the dry run without one; any"
+    echo "[dry-run] Continuing dry run without one -- any 'would re-encrypt' output below"
-    echo "[dry-run] 'would re-encrypt' output below couldn't actually run for real yet."
+    echo "[dry-run] couldn't actually run for real until a key is present."
     return
   fi
 
-  echo "No sops age decryption key found (checked \$SOPS_AGE_KEY, \$SOPS_AGE_KEY_FILE, ${key_file})."
+  cat >&2 <<EOF
-  echo "Generating a new one at ${key_file}..."
+No sops age decryption key found (checked \$SOPS_AGE_KEY, \$SOPS_AGE_KEY_FILE, ${key_file}).
-  mkdir -p "$(dirname "$key_file")"
-  nix-shell "${NIX_OPTS[@]}" -p age --run "age-keygen -o '${key_file}'" 2>&1 | grep -v "^Public key:" || true
-  local new_pub
-  new_pub="$(age_pubkey_from_identity_file "$key_file")"
 
-  cat <<EOF
+Place your admin age private key at ${key_file}, or set SOPS_AGE_KEY (inline
+key) or SOPS_AGE_KEY_FILE (path to a different key file) and re-run.
 
-A brand-new age key was just generated -- it cannot decrypt anything that
+If the key is truly missing (not just mislocated), this is a manual recovery
-already exists in secrets/*.yaml, since nothing was ever encrypted for it.
+situation -- generating a brand-new admin key won't help, since it cannot
-That trust can't be bootstrapped automatically (nobody can decrypt a file
+decrypt anything already encrypted for the old one.  Each secrets/*.yaml is
-for a recipient that didn't exist when it was last encrypted).
+also encrypted for its respective host key(s), so a running deployed host can
-
+still decrypt what it needs -- but the admin key is required for re-encryption
-To actually use this key:
+(e.g. adding new recipients via sops updatekeys).
-  1. Have someone who currently CAN decrypt replace the &admin entry in
-     .sops.yaml with this public key:
-       ${new_pub}
-  2. They re-encrypt every secrets/*.yaml:
-       sops updatekeys --yes secrets/common.yaml
-       sops updatekeys --yes secrets/nix-cache.yaml
-       sops updatekeys --yes secrets/server.yaml
-  3. Re-run this script.
-
-Exiting without making any other changes.
 EOF
   exit 1
 }

secrets/ha-corosync-authkey

@@ -0,0 +1 @@
+STUB: run cluster-init.sh to generate, then: sops -e --input-type binary /etc/corosync/authkey > secrets/ha-corosync-authkey

secrets/ha-server-1.yaml

@@ -0,0 +1,6 @@
+# STUB — not yet encrypted with sops.
+# Bootstrap:
+#   bash scripts/secrets/sync-host-keys.sh proxmox-ha-server-1
+#   sops updatekeys secrets/common.yaml  (allows ha-server-1 to decrypt shared secrets)
+#   sops secrets/ha-server-1.yaml        (create with: beszel-token)
+beszel-token: REPLACE

secrets/ha-server-2.yaml

@@ -0,0 +1,6 @@
+# STUB — not yet encrypted with sops.
+# Bootstrap:
+#   bash scripts/secrets/sync-host-keys.sh proxmox-ha-server-2
+#   sops updatekeys secrets/common.yaml  (allows ha-server-2 to decrypt shared secrets)
+#   sops secrets/ha-server-2.yaml        (create with: beszel-token)
+beszel-token: REPLACE

secrets/tailscale-router.yaml

@@ -0,0 +1,43 @@
+beszel-token: ENC[AES256_GCM,data:qg2eb7Eq/w7vIuiiDeK/s6h9nmHWuooyrd93GvdLD2hGOvka,iv:FWRBQTIY9PzV2V8J7z5ftEb5dvjN/S0hP0H9JV+Uuq8=,tag:WP7DHOu4elfqdLoW7vo9qg==,type:str]
+sops:
+    age:
+        - enc: |
+            -----BEGIN AGE ENCRYPTED FILE-----
+            YWdlLWVuY3J5cHRpb24ub3JnL3YxCi0+IFgyNTUxOSBCZmozK3VjcENGb3l2MDNj
+            QVE3VVI3Z3R0MjduSlJGL3BFWkg5NHZmZms0CkNBTWZKZFJVNkJMZ0tpNXhXWW5q
+            WndvbmhuOVh5ZHhJMmJwenh0M2dEVlUKLS0tIGpXQ3dhRXVWUDZzdGg4RHN4V3p0
+            WDZnK1lMUHVIS09hRUEzVHRHQ0EwcWsKEd5NtstHSPH8okXcFLW7BId5xPGmgyC3
+            dAodIGkLBlJlaL1bZuRy6Vaac+mJZL5KJYD/o1kYSXRiHJfNpaulTg==
+            -----END AGE ENCRYPTED FILE-----
+          recipient: age1njap586hc0q43kr03g6c8eqhdsmk8zcafkl3f83xwlc2gqhlmfgs4tmwad
+        - enc: |
+            -----BEGIN AGE ENCRYPTED FILE-----
+            YWdlLWVuY3J5cHRpb24ub3JnL3YxCi0+IFgyNTUxOSBXWWhYOFU3VHZ3S1VuRTF2
+            QlZvMkJUNDZmT1RDWWZsMkxnZjJldjdrRG44CkU0NEJTR0NhQUpjMEY4RXBLSVBJ
+            bDNtZ1UrRDNUbzA0NDJ1cWJKdlBBMWMKLS0tIE9vTmEzTUxpODF4SG1MQ1czM1N5
+            dVgrMERMRXJwVmhtMldTZUlHNUJVTEkKzUf9cETbo0K9bCPkrlbUkgkhRIPvjCHR
+            GFVwgb2fcaOzZHcDwlLDLaP6HbRPtzJfVbLpQ1XU0N/xijSr9ENB6g==
+            -----END AGE ENCRYPTED FILE-----
+          recipient: age1f7usptjx9rv4rxauasve200gxtdt9jkqhhdqstlf20wvlm7u75rsjfw50m
+        - enc: |
+            -----BEGIN AGE ENCRYPTED FILE-----
+            YWdlLWVuY3J5cHRpb24ub3JnL3YxCi0+IFgyNTUxOSBlOVdWN0tHNWZmaWlWQ2hJ
+            MVFYd0pMNmJTdTlJQ3h1c0dzbDE3U2VqdzNzCitCcFdhbjRPYnl3Q214WnRxa2Jq
+            eTFnc2VBcXlGcTVYdWpWTFZiWXBMUDAKLS0tIC9YdVkrTU5EeXNlNzliLzBlYUR2
+            Qk1ITW9tMElzUHF4MVJyWS9uRGs1QncKj8OgxWYabf8MccWD6zYLxl+FvYPAvLCW
+            p/FJ8uMa9YWp0RjICvWIwyO0gK7oaxTh/JzfBjABngEZnZazNYP7Uw==
+            -----END AGE ENCRYPTED FILE-----
+          recipient: age1k7d2du5mejsmv5rzavm4xwgpthqvcfsehduquv28nzs53zppa3kqngfxq2
+        - enc: |
+            -----BEGIN AGE ENCRYPTED FILE-----
+            YWdlLWVuY3J5cHRpb24ub3JnL3YxCi0+IFgyNTUxOSBsbzFKckFQVDg0VEpJMFlv
+            WSs1ODVoYlZZK3YrSXlMU1hIQjdLTUhNdHhJCkRnbHY5Z2NGQVMvOHV4R2hTd0xu
+            LzdheXdZVVBsTjhVOU8zZGFKN0lPWjQKLS0tIFdaaG5SeElDN3YvMjFxdHVLWlNX
+            M1BQQmtJTUdsK2x5TEo5TmNrMEZidkEKhi4jNKhrkBAPhJoeYNg1d60LQTqmJt+G
+            Z/2X5uWzh12zF5keL8cNHH6sD0z0MOChtd4UF37yr+lb/LFJkia6LQ==
+            -----END AGE ENCRYPTED FILE-----
+          recipient: age1zhfyuzlq40reuqlr34gf77852nhs3t6mqfzrqmas8z6sxk7tcfhsungrm0
+    lastmodified: "2026-07-26T01:12:11Z"
+    mac: ENC[AES256_GCM,data:0r68y6XmdiW7p/h9QmTR5h37TxKp+0wF44LP8QfG6cGP0tvaasuvv+V9EC3msnREYjwJhlnHjRd8hJ31HGVkOHbYWI4V5cF4ZnaiGgrZJhsGcJvOJSroAvQwev+vIlXZpAPgZz3Z35FDvOpgb71l5c0so0l4EJhc70JOh2FxhvQ=,iv:Or6l/v/E01+9wtZwyl+pCttiuReR6V0kmBzE9KhsfgQ=,tag:cAEUgj/9tKDToKFr00g5UQ==,type:str]
+    unencrypted_suffix: _unencrypted
+    version: 3.13.2

variables.nix

@@ -4,8 +4,19 @@
   homeDomain = "sweet.home"; # base LAN domain for service subdomains (pve., docker.)
   tailnetDomain = "tail13f623.ts.net"; # Tailscale MagicDNS suffix
   lanCidr = "192.168.2.0/24"; # LAN subnet
-  pxeServerIp = "192.168.2.247"; # pxe-boot host's LAN IP
+  lanGateway = "192.168.2.254"; # LAN default gateway (router)
-  pbsIp = "192.168.2.108"; # Proxmox Backup Server LAN IP
+  lanPrefixLength = 24; # LAN subnet prefix length (/24 = 255.255.255.0)
+  lxcLanInterface = "eth0"; # LAN NIC name in LXC containers (set by Proxmox --net0 name=eth0)
+  vmLanInterface = "ens18"; # LAN NIC name in Proxmox VMs (virtio, first NIC)
+  vmStorageInterface = "ens19"; # storage NIC name in HA server VMs (virtio, second NIC on vmbr1)
+  pxeServerIp = "192.168.2.223"; # pxe-boot LXC container LAN IP
+  nixCacheIp = "192.168.2.224"; # nix-cache LXC container LAN IP
+  tailscaleRouterIp = "192.168.2.222"; # tailscale-router LXC container LAN IP
+  torRelayIp = "192.168.2.221"; # tor-relay LXC container LAN IP
+  serverIp = "192.168.2.226"; # server (NFS/ZFS) Proxmox VM LAN IP
+  dockerIp = "192.168.2.225"; # docker Proxmox VM LAN IP
+  pbsIp = "192.168.2.244"; # Proxmox Backup Server LAN IP (not NixOS-managed)
+  domainControllerIp = "192.168.2.253"; # FreeIPA domain controller / primary DNS (not NixOS-managed)
 
   # Cross-host references (LAN hostnames/users other hosts reach over the network)
   nixCacheHost = "nix-cache"; # substituter/remote-builder hostname
@@ -68,6 +79,24 @@
   # one-line change.
   primaryUser = "nixos";
 
+  # HA file server cluster
+  # LAN IPs (vmbr0 / ens18) — client-facing: iSCSI initiators, NFS, management.
+  # Storage IPs (vmbr1 / ens19) — isolated internal bridge, used for DRBD
+  # replication and Corosync heartbeat only; never leaves pve1.
+  # haServerVip: floating virtual IP managed by Pacemaker's IPaddr2 resource;
+  # NFS and iSCSI clients connect here regardless of which node is Active.
+  haServer1Host = "ha-server-1";
+  haServer2Host = "ha-server-2";
+  haServer1Ip = "192.168.2.228"; # LAN IP, node 1
+  haServer2Ip = "192.168.2.227"; # LAN IP, node 2
+  haServerVip = "192.168.2.229"; # floating VIP (Pacemaker IPaddr2)
+  haServer1StorageIp = "192.168.4.228"; # storage-net IP, node 1 (vmbr1 / ens19)
+  haServer2StorageIp = "192.168.4.227"; # storage-net IP, node 2 (vmbr1 / ens19)
+  haStorageCidr = "192.168.4.0/29"; # storage subnet — internal to pve1 only
+  haStoragePrefixLength = 29; # storage subnet prefix length (/29)
+  haStorageRoot = "/srv/ha-data"; # XFS-over-DRBD mount point on the Active node
+  haIscsiIqn = "iqn.2026-01.home.sweet:ha-storage";
+
   # Storage
   storageRoot = "/tank"; # ZFS pool root on `server`
 
@@ -83,6 +112,7 @@
   # dataset or moving where it's mounted only needs changing it here — the
   # export and every client reference follow automatically.
   nfsShares = {
+    options = "(rw,sync,no_subtree_check,no_root_squash)";
     dockerConfig = {
       subpath = "docker/config";
       mountpoint = "/mnt/docker/config";
@@ -103,6 +133,18 @@
       subpath = "raspi/volumes";
       mountpoint = "/mnt/raspi-backup";
     };
+    proxmoxIsos = {
+      subpath = "proxmox/iso";
+      mountpoint = "/mnt/iso";
+    };
+    proxmoxLxcImages = {
+      subpath = "proxmox/lxc";
+      mountpoint = "/mnt/lxc";
+    };
+    pxebootImages = {
+      subpath = "pxe-boot/images";
+      mountpoint = "/mnt/pxe-images";
+    };
   };
 
   # The Raspberry Pi's own NFS export — not under storageRoot/nfsServerHost,
@@ -129,10 +171,25 @@
     # (modules/build-types/pxe-boot.nix).
     pxeBootTftp = 69;
 
-    # `server`'s NFS exports need both the portmapper (rpcbind) and the
+    # `server`'s NFS exports: portmapper (rpcbind), NFS data, and the
-    # NFS data port itself opened (modules/build-types/server.nix).
+    # mountd RPC service (used by showmount/NFSv3 mount protocol).
+    # Mountd listens on a fixed port so the firewall can whitelist it
+    # explicitly rather than opening all of rpcbind's dynamic range.
+    # All three need both TCP and UDP (modules/build-types/server.nix and
+    # modules/build-types/ha-server.nix).
     nfsRpcbind = 111;
     nfsd = 2049;
+    nfsMountd = 20048;
+
+    # HA cluster ports opened on ha-server-1 and ha-server-2
+    # (modules/build-types/ha-server.nix / modules/ha/cluster-config.nix).
+    haServerDrbd = 7789; # DRBD replication (TCP)
+    haServerIscsi = 3260; # iSCSI target (TCP)
+    haServerCorosync1 = 5404; # Corosync totem ring (UDP)
+    haServerCorosync2 = 5405; # Corosync totem ring (UDP)
+    haServerCorosyncCrypto = 5407; # Corosync crypto sync (UDP)
+    haServerPacemakerRemoted = 3121; # pacemaker-remoted (TCP)
+    haServerPcsd = 2224; # pcsd cluster daemon (TCP)
 
     # Opened on the docker host's firewall for the Traefik-fronted
     # container stack (docker-compose config lives in the separate