fix(ha): per-node DRBD disk assignment (NODE2 data disk is /dev/sda not /dev/sdb)

Proxmox VM disk ordering differs between the two HA nodes:
  ha-server-1: sda=OS (50G), sdb=DRBD data (32G)
  ha-server-2: sda=DRBD data (32G), sdb=OS (50G)

The DRBD resource config was using a shared disk=/dev/sdb which targeted
the OS disk on ha-server-2, causing drbdmeta and drbdadm up to operate
on the mounted root filesystem (hence "Device or resource busy").

Changes:
- variables.nix: add haServer1DrbdDisk/haServer2DrbdDisk
- cluster-config.nix: move volume block inside per-host on{} sections
  so each node uses the correct backing disk
- cluster-init.sh: use NODE1_DRBD_DISK/NODE2_DRBD_DISK variables;
  add runtime check that patches /etc/drbd.d/*.res on the running nodes
  if the deployed config points to the wrong disk (workaround for VMs
  built before this fix; redeploy with --force-rebuild to make permanent)

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

variables.nix

@@ -108,6 +108,11 @@
   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";
+  # DRBD backing device on each node. Disk ordering can differ between Proxmox
+  # VMs depending on the order disks were added; these must match the actual
+  # block device that is NOT the OS disk on each node (verify with lsblk).
+  haServer1DrbdDisk = "/dev/sdb"; # data disk on ha-server-1 (OS disk is /dev/sda)
+  haServer2DrbdDisk = "/dev/sda"; # data disk on ha-server-2 (OS disk is /dev/sdb)
 
   # Storage
   storageRoot = "/tank"; # ZFS pool root on `server`