wangjia
3d5bac66b4
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>
82 lines
3.2 KiB
Go
82 lines
3.2 KiB
Go
package provision
|
|
|
|
import (
|
|
"context"
|
|
"time"
|
|
)
|
|
|
|
// ReplaceStep enumerates the resumable steps of a Replace orchestration.
|
|
// Persisted in replacements.step so a crashed process can continue without
|
|
// repeating a boot or a destroy.
|
|
type ReplaceStep string
|
|
|
|
const (
|
|
StepOpenNew ReplaceStep = "open_new" // create the replacement node
|
|
StepProbing ReplaceStep = "probing" // wait for self-register + probe
|
|
StepNewUp ReplaceStep = "new_up" // promote new node, bump version
|
|
StepDraining ReplaceStep = "draining" // drain the old node
|
|
StepDestroyOld ReplaceStep = "destroy_old" // destroy old node + release IP
|
|
StepDone ReplaceStep = "done"
|
|
)
|
|
|
|
// ReplaceStatus is the terminal/running state of a replacement record.
|
|
type ReplaceStatus string
|
|
|
|
const (
|
|
ReplaceRunning ReplaceStatus = "running"
|
|
ReplaceDone ReplaceStatus = "done"
|
|
ReplaceFailed ReplaceStatus = "failed"
|
|
)
|
|
|
|
// Replacement mirrors a replacements row (crash-recoverable orchestration).
|
|
type Replacement struct {
|
|
UUID string
|
|
OldNodeID int64
|
|
NewNodeID int64 // 0 until the new node is created
|
|
Pool Pool
|
|
Status ReplaceStatus
|
|
Step ReplaceStep
|
|
CreatedAt time.Time
|
|
UpdatedAt time.Time
|
|
}
|
|
|
|
// Store is the persistence boundary for the provision package. Both the MySQL
|
|
// implementation (mysqlStore) and the in-memory test fake satisfy it, keeping
|
|
// the service / orchestration logic database-agnostic and unit-testable.
|
|
type Store interface {
|
|
// --- nodes ---
|
|
InsertNode(ctx context.Context, n *Node) (int64, error)
|
|
GetNode(ctx context.Context, id int64) (*Node, error)
|
|
GetNodeByUUID(ctx context.Context, uuid string) (*Node, error)
|
|
UpdateNodeStatus(ctx context.Context, id int64, status Status) error
|
|
UpdateNodeEndpoint(ctx context.Context, id int64, endpoint string) error
|
|
// SetNodeInstance records the vendor instance ID and IP-derived endpoint.
|
|
SetNodeInstance(ctx context.Context, id int64, instanceID, endpoint string) error
|
|
SetNodeWeight(ctx context.Context, id int64, weight int) error
|
|
// ListNodesByPool returns nodes whose provider belongs to pool, optionally
|
|
// filtered to a single status (empty = all statuses).
|
|
ListNodesByPool(ctx context.Context, pool Pool, status Status) ([]*Node, error)
|
|
|
|
// --- providers ---
|
|
ListProviders(ctx context.Context, pool Pool) ([]*Provider, error)
|
|
GetProvider(ctx context.Context, id int64) (*Provider, error)
|
|
|
|
// --- events / audit / directory ---
|
|
WriteNodeEvent(ctx context.Context, nodeID int64, event Event, detailJSON string) error
|
|
WriteAuditLog(ctx context.Context, actor, action, target, metaJSON string) error
|
|
BumpDirectoryVersion(ctx context.Context) (int64, error)
|
|
|
|
// --- idempotency ---
|
|
// LookupIdempotency returns the node UUID previously bound to key, or
|
|
// (\"\", false, nil) if unseen.
|
|
LookupIdempotency(ctx context.Context, key string) (string, bool, error)
|
|
// SaveIdempotency binds key→nodeUUID. It is a no-op if the key already
|
|
// exists (first write wins).
|
|
SaveIdempotency(ctx context.Context, key, nodeUUID string) error
|
|
|
|
// --- replacements (crash recovery) ---
|
|
CreateReplacement(ctx context.Context, r *Replacement) error
|
|
GetReplacement(ctx context.Context, uuid string) (*Replacement, error)
|
|
UpdateReplacement(ctx context.Context, r *Replacement) error
|
|
}
|