feat(provision): 弹性节点基建 Terraform + 一键更换 (tsk_6u0FxmbC7Yeq)

IaC 面 (infra/) 与控制面 (server/internal/provision) 双产出,落地 doc/04 §4
「节点是牲口」弹性拓扑与 make-before-break 一键更换。

server/internal/provision:
- CloudAdapter 适配层 + Registry(首发 vultr 消耗品池 / hetzner 精品池各一);
  厂商凭证仅从 PROVISION_<VENDOR>_* env 注入,不入库不入 git。
- ProvisionService:CreateNode(幂等键重放不重复开机)、DestroyNode(幂等)、
  RotateIP(换 IP 不换机 + version bump)、ListProviders。
- Replace 一键更换:先建后拆,新机 up 先于旧机 draining(容量不下降),
  replacement_uuid 幂等键 + replacements 表分步记录,崩溃可续跑不重复。
- RotatePool:池内滚动轮换,并发度 1–2。
- cmd/nodectl CLI:create/destroy/rotate-ip/replace/rotate-pool/providers。
- 单测(mock 厂商 API + 内存 Store):幂等重放、make-before-break 时序断言、
  开机失败/探活超时→destroyed+失败计数+告警钩子、崩溃续跑、RotatePool。

infra/:
- terraform/:探针机 + 控制面基线模块化(probe / control-plane)+ README,
  低频基线进 state,节点不进 Terraform。
- cloud-init/node.yaml.tmpl:节点引导模板(注入一次性 bootstrap token,task #5)。
- identity-isolation.md:身份隔离登记表(doc/06 §2 红线,无任何凭证)。

migrations/000008:nodes 增 provider_instance_id/elastic_ip_id、node_events
增 ip_rotated、provision_idempotency / replacements 表(附加式,不动现网)。

红线:仅面向新厂商池,绝不纳管现网生产 EC2(deploy/ marzban)。

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
This commit is contained in:
wangjia
2026-06-13 14:23:39 +08:00
parent 787151245e
commit 3d5bac66b4
39 changed files with 3884 additions and 0 deletions
+31
View File
@@ -0,0 +1,31 @@
// Package provision owns the elastic-node control plane: it turns the
// "nodes are cattle, not pets" principle (doc/04 §2) into running code.
//
// Responsibility split (doc/04 §4):
//
// - Terraform (infra/terraform) manages low-frequency baseline resources:
// probe machines and the control-plane environment. Those live in
// Terraform state.
// - This package drives high-frequency, minute-scale node lifecycle through
// vendor APIs. Disposable nodes are NOT in Terraform state.
//
// The package exposes a ProvisionService with idempotent operations:
//
// - CreateNode — insert nodes(status=provisioning) → vendor API boot →
// render cloud-init (injecting a one-time bootstrap token, task #5) →
// return. The agent self-registers (task #6) and flips the node to up.
// - DestroyNode — vendor destroy + IP release + nodes→destroyed.
// - RotateIP — swap the elastic IP without re-creating the machine
// (change IP, keep the box), then bump the directory version.
// - Replace — make-before-break one-click replacement: bring a fresh
// node up BEFORE draining/destroying the old one, so capacity never dips.
// - RotatePool — rolling Replace across a whole pool, concurrency 12.
//
// Vendor adapters live in providers/ behind the CloudAdapter interface.
// Vendor credentials are injected ONLY from independent secrets (env/file) —
// never stored in the database and never committed to git. The providers table
// holds just name/api_kind/regions/pool/enabled.
//
// Hard red line (doc/06 §2): every operation here targets the new vendor pools.
// It MUST NOT ever manage the production EC2 host (deploy/ marzban machine).
package provision