DmitryandClaude Sonnet 5
a3fe8031fe
Blue-green: a new container is created beside the old one, data is copied, the
IP is moved onto it, and the old container is kept stopped as rollback for at
least a week. Keeping the IP means only the VMID changes, and its consumers
(backup jobs, backup audit) already derive it from the registry.
Batch 1 (2026-09-02): emergency-bot 148->151, docker-test 145->152,
gitea 141->153, vaultwarden 140->154, monitoring 146->155, gyro 150->156,
grimmory 149->157.
Batch 2 (2026-09-03): adguard 144->158, mihomo 143->159, ovpn-mini 132->160.
All migratable LXC are now provisioner: tofu. hermes-ai (frozen) and pbs stay.
- tofu/services.tf + tofu/svc-*.tf: one resource per service, reproducing the
pct-config etalon. /dev/fuse -> features.fuse; /dev/net/tun ->
device_passthrough (first live use on mihomo and ovpn-mini); gitea bind mount
-> datastore volume (data finally reaches PBS); console { type = "shell" }
declared explicitly (provider tracks cmode there).
- services.yml: vmid + provisioner: tofu for every migrated service; features
strings and device notes updated to the tofu representation; also drops the
memoir-bot entry and adds homelab_reverse_proxy_image/_unit.
- pve-*.yml: configuration play target is `{{ pve_config_target | default(...) }}`
so it can run against <name>-new on a temp address (a bare --limit zeroes the
play instead of retargeting it). Container-creation plays are gated behind
`provisioner != 'tofu'` / `pve_provisioning_enabled` (meta: end_play), so a
stray run cannot pct start a stopped OLD VMID on a live IP. Override for
intentional legacy rollback: -e pve_<svc>_legacy_provisioning_enabled=true.
- ssh_config: drop memoir-bot; ovpn-mini gets ProxyJump none (a jump via ru-vps
would route through the very tunnel ovpn-mini terminates).
- gyro.yml / uptime-kuma.yml: same pve_config_target override.
- roles/uptime_kuma: only freeze homelab-monitoring when the unit actually
exists (a fresh blue-green container never had it).
- offsite-restic-yadisk.yml: the gitea restic profile now runs inside the LXC
(hosts: gitea), since the bind-mount host path is gone after the volume move;
lost+found excluded (unreadable in an unprivileged LXC, restic exit 3).
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_012uoq5AVK8mkBgg83Mq6o5V
140 lines
5.3 KiB
Terraform
140 lines
5.3 KiB
Terraform
# ============================================================================
|
||
# Сервис №9 по плану (docs/ai/migration-tofu.md, раздел 5): mihomo.
|
||
# Локальный прокси + MetaCubeXD UI. На него завязаны egress hermes-ai
|
||
# (bash_config_proxy_* в hosts.yml), внешние пробы Uptime Kuma, emergency-бот
|
||
# и уведомления gyro — простой затрагивает всё это разом.
|
||
#
|
||
# OLD vmid 143, узел mini-pc, боевой адрес 192.168.1.27.
|
||
# Эталон снят 2026-09-02: ssh mini-pc sudo pct config 143 (дословно совпадает
|
||
# с /etc/pve/lxc/143.conf):
|
||
#
|
||
# arch: amd64
|
||
# cmode: shell
|
||
# cores: 1
|
||
# features: nesting=1,keyctl=1
|
||
# hostname: mihomo
|
||
# memory: 512
|
||
# nameserver: 1.1.1.1
|
||
# net0: name=eth0,bridge=vmbr0,firewall=1,gw=192.168.1.1,ip=192.168.1.27/24,type=veth
|
||
# onboot: 1
|
||
# ostype: debian
|
||
# rootfs: local-lvm:vm-143-disk-0,size=8G
|
||
# startup: order=70
|
||
# swap: 512
|
||
# unprivileged: 1
|
||
# lxc.cgroup2.devices.allow: c 10:229 rwm # /dev/fuse
|
||
# lxc.mount.entry: /dev/fuse dev/fuse none bind,create=file
|
||
# lxc.cgroup2.devices.allow: c 10:200 rwm # /dev/net/tun
|
||
# lxc.mount.entry: /dev/net/tun dev/net/tun none bind,create=file
|
||
#
|
||
# vm_id 159 — новый VMID для blue-green переезда (назначен, не менять).
|
||
# TMPIP 192.168.1.18 используется только на шаге 4.3. После шага 4.5.15 адрес
|
||
# меняется на боевой 192.168.1.27, OLD (143) останавливается и остаётся
|
||
# откатом минимум неделю.
|
||
#
|
||
# --- Два устройства, ДВА разных механизма -----------------------------------
|
||
# /dev/fuse -> штатный флаг features.fuse (Docker внутри на fuse-overlayfs).
|
||
# /dev/net/tun -> блок device_passthrough (dev0: в PVE 8.2+). mihomo.service
|
||
# запускает контейнер с `--device /dev/net/tun` для TUN-режима.
|
||
# Это первый боевой сервис, использующий device_passthrough: на пилоте VMID
|
||
# 199 механизм проверен (`dev0: deny-write=0,path=/dev/net/tun,...`, повторный
|
||
# plan -> No changes), см. tofu/pilot.tf.example и tofu/README.md. Оба требуют
|
||
# root@pam; в обычном токенном режиме Proxmox отвечает 403.
|
||
# Следствие: pct config покажет `features: fuse=1,...` и `dev0: path=/dev/net/tun`
|
||
# вместо сырых lxc.* строк эталона — тот же эффект штатным механизмом.
|
||
#
|
||
# --- Данные ----------------------------------------------------------------
|
||
# /opt/mihomo (~80 КБ): config/config.yaml (реальный конфиг прокси),
|
||
# config/cache.db, config/providers/main.yaml. Переносится целиком на шаге
|
||
# 4.4. Задача "Install default mihomo config if missing" в pve-mihomo.yml
|
||
# идёт с force: false, поэтому перенесённый конфиг не перетирается.
|
||
# ============================================================================
|
||
|
||
resource "proxmox_virtual_environment_container" "mihomo" {
|
||
node_name = "mini-pc"
|
||
vm_id = 159
|
||
unprivileged = true
|
||
start_on_boot = true
|
||
started = true
|
||
tags = ["tofu"]
|
||
|
||
initialization {
|
||
hostname = "mihomo"
|
||
|
||
ip_config {
|
||
ipv4 {
|
||
address = "192.168.1.27/24"
|
||
gateway = "192.168.1.1"
|
||
}
|
||
}
|
||
|
||
dns {
|
||
servers = ["1.1.1.1"]
|
||
}
|
||
|
||
user_account {
|
||
keys = [trimspace(file(pathexpand("~/.ssh/id_ed25519_homelab.pub")))]
|
||
}
|
||
}
|
||
|
||
operating_system {
|
||
template_file_id = "local:vztmpl/debian-13-standard_13.1-2_amd64.tar.zst"
|
||
type = "debian"
|
||
}
|
||
|
||
cpu {
|
||
cores = 1
|
||
}
|
||
|
||
memory {
|
||
dedicated = 512
|
||
swap = 512
|
||
}
|
||
|
||
disk {
|
||
datastore_id = "local-lvm"
|
||
size = 8
|
||
}
|
||
|
||
# Эталон: features: nesting=1,keyctl=1 плюс /dev/fuse через сырые lxc.*
|
||
# строки. fuse=1 — штатная замена этого обхода (см. комментарий выше).
|
||
features {
|
||
nesting = true
|
||
keyctl = true
|
||
fuse = true
|
||
}
|
||
|
||
# /dev/net/tun: в эталоне проброшен сырыми lxc.cgroup2.devices.allow +
|
||
# lxc.mount.entry. Штатная декларация — device_passthrough (dev0:).
|
||
device_passthrough {
|
||
path = "/dev/net/tun"
|
||
}
|
||
|
||
# roles/pve_lxc всем контейнерам ставит cmode=shell (pve_lxc_cmode), провайдер
|
||
# это отслеживает через блок console.type — без него на следующем apply
|
||
# откатит на дефолт Proxmox tty. Обнаружено 2026-09-02 на emergency-bot,
|
||
# см. tofu/README.md.
|
||
console {
|
||
type = "shell"
|
||
}
|
||
|
||
network_interface {
|
||
name = "eth0"
|
||
bridge = "vmbr0"
|
||
firewall = true
|
||
}
|
||
|
||
startup {
|
||
order = 70
|
||
}
|
||
}
|
||
|
||
output "mihomo" {
|
||
description = "Что проверять на узле после apply"
|
||
value = {
|
||
vmid = proxmox_virtual_environment_container.mihomo.vm_id
|
||
node = proxmox_virtual_environment_container.mihomo.node_name
|
||
verify = "ssh mini-pc sudo pct config ${proxmox_virtual_environment_container.mihomo.vm_id}"
|
||
}
|
||
}
|