admin
c70d5d860e
背景：把平台从「单机单进程」改造成「1 组 Manager + N 台 Worker + 共享归属状态」，
硬约束 = 全程兼容单例模式（deployMode 默认 local；生产切换前 47 一行未动）。
主要改动
1) 数据模型 v7（SQLite 与 PG 两方言同步）：新增 dsh_hosts 注册表 +
dsh_instances.{host_id,epoch,heartbeat_at,lease_until}；claimInstance 原子抢占
（UPDATE … WHERE host_id IS NULL OR lease_until < now）+ pinInstanceHost 钉住归属。
2) 租约与 fencing：src/supervisor/lease.ts（acquire/renew/release + stillHolder 判据 +
ttl > 2×renew 硬校验）；心跳里续租，失权即向 worker 下发更高 epoch（self-fencing）。
⚠️ release 只清租约（lease_until），**保留 host_id** —— host_id 是「用户数据在哪台」的锚点。
3) Worker agent（src/worker/agent.ts，子命令 dshs worker）：实例生命周期 + 文件面 /fs/*
+ 幂等键（operationId）+ 鉴权（timingSafeEqual）；Worker 不写控制面数据
（apiKey/uid 由 Manager 随 launch 投递，R5 收窄）。
4) 远端 Spawner + LeasedSpawner：按 host 路由（**粘性优先**：有历史归属且那台 up 就留在原地，
否则按容量选最空的）+ 容量准入 + deployMode=cluster 装配（systemd drop-in，可回滚）。
5) bwrap 修正：**所有挂载点的中间目录统一前置 + 去重 + 由外到内**（「就近创建」会在嵌套前缀下
遮掉已绑挂载点 ⇒ bwrap: Can't chdir）；且**只能用 --tmpfs**，用 --perms 会让 47 的
bwrap 0.4.0 直接拒启动（沙箱全挂）。
6) 跨机隧道 src/worker/tunnel.ts：SSH ControlMaster + 动态 -R 转发；**自愈由 agent 本地
20s 定时器驱动**（不能只放 /healthz —— 心跳本身经隧道进来，断了就没人触发它）。
7) 文件面按归属路由（RemoteUserFs）：实例与文件必须落在同一台机器，否则实例看不到自己的文件。
8) 观测面：dshs doctor / dshs cluster status。
验证（本次均已实跑）
- test/lease.test.mjs：SQLite 10/10 == PG 10/10
- 组件级端到端 5 个：verify-cluster-{agent,lease,fs,migrate,live}.mjs
- 真跨机（47 Manager / 106 Worker，跨云 + 反向隧道）verify-cluster-cross.mjs 九步全绿
- 域名形态访问 verify-cluster-domain.mjs（<user>.域名 → Manager → 远端实例；越权 403）
- 冒烟 scripts/smoke-*：6/8，失败项与改动前基线完全相同（无回归）
- 生产切换与回滚剧本见 dsh-server-docs/交接单/T08-集群化落地-兼容单例模式.md §16
80 lines
5.1 KiB
Bash
80 lines
5.1 KiB
Bash
#!/usr/bin/env bash
|
||
# 最终验证（在 47 上跑）：清污染 → 重置锚点 → 重验两条路径
|
||
# ① 既有用户 guest：留 w-47 + 工作区有历史数据 + 实例页正常
|
||
# ② 新用户：落 w-106 + **文件真的写到 106 的盘** + 实例页正常
|
||
set -uo pipefail
|
||
export PGPASSWORD=dshs_cluster_2026
|
||
Q() { /usr/bin/psql -h 127.0.0.1 -p 15432 -U dshs -d dshs -tAc "$1"; }
|
||
M=http://127.0.0.1:3080
|
||
DOMAIN=alotbuy.com
|
||
T47=dshs-worker-47-c4b7e19f
|
||
T106=dshs-worker-7f3a91c05e
|
||
SSH106="ssh -n -i /root/.ssh/dshworker_ed25519 -o BatchMode=yes -o StrictHostKeyChecking=accept-new [email protected]"
|
||
|
||
mksess() {
|
||
local u="$1" T H
|
||
T=$(openssl rand -hex 32); H=$(printf '%s' "$T" | sha256sum | cut -d' ' -f1)
|
||
Q "insert into sessions (token_hash,user_id,created_at,expires_at,ip,user_agent)
|
||
select '$H', id, (extract(epoch from now())*1000)::bigint, (extract(epoch from now())*1000)::bigint+1800000, '127.0.0.1','switch-verify' from users where username='$u'" >/dev/null
|
||
printf '%s' "$T"
|
||
}
|
||
owner() { Q "select coalesce(i.host_id,'NULL')||' epoch='||coalesce(i.epoch,0) from users u left join dsh_instances i on i.user_id=u.id where u.username='$1'"; }
|
||
agent() { curl -s -m 6 -H "x-dsh-agent-token: $2" "http://127.0.0.1:$1/healthz" | grep -o '"instances":[0-9]*'; }
|
||
isapp() { grep -q '<base href=' <<<"$1" && echo "✓实例页" || echo "✗非实例页"; }
|
||
|
||
echo "########## 0) 清污染：停所有实例 + 删测试用户 ##########"
|
||
systemctl restart dshs-worker; sleep 4
|
||
$SSH106 'systemctl restart dshs-worker' >/dev/null 2>&1; sleep 4
|
||
echo " 重启后 w-47=$(agent 19100 $T47) w-106=$(agent 19000 $T106)"
|
||
AT=$(mksess admin); AC="sid=$AT"
|
||
for u in $(Q "select id from users where username like 'switchtest%' or username like 'swtest%'"); do
|
||
echo " 删测试用户 $u → $(curl -s -m 20 -X DELETE -b "$AC" "$M/api/admin/users/$u" -o /dev/null -w '%{http_code}')"
|
||
done
|
||
echo " 剩余用户: $(Q "select string_agg(username,', ') from users")"
|
||
|
||
echo "########## 1) 重置 guest 锚点（host_id=w-47） ##########"
|
||
Q "update dsh_instances set host_id='w-47', epoch=0, lease_until=0, status='stopped'
|
||
where user_id=(select id from users where username='guest')" >/dev/null
|
||
echo " $(owner guest)"
|
||
|
||
echo
|
||
echo "########## 2) 路径①：既有用户 guest ##########"
|
||
GT=$(mksess guest); GC="sid=$GT"
|
||
E=$(curl -s -m 60 -X POST -b "$GC" -H 'content-type: application/json' -d '{}' "$M/api/dsh/enter")
|
||
sleep 3
|
||
echo " 归属: $(owner guest) ← 期望 w-47"
|
||
echo " w-47=$(agent 19100 $T47) w-106=$(agent 19000 $T106)"
|
||
echo " 工作区条目: $(curl -s -m 10 -b "$GC" "$M/api/desktop/tree" | grep -o '"name":"[^"]*"' | head -4 | tr '\n' ' ')"
|
||
PAGE=$(curl -s -m 25 -L -b "$GC" -H "Host: guest.$DOMAIN" "$M/" | head -c 300)
|
||
echo " 实例页: $(isapp "$PAGE")"
|
||
|
||
echo
|
||
echo "########## 3) 路径②：新用户 ##########"
|
||
NU="swtest2$(date +%H%M%S)"
|
||
echo " register=$(curl -s -m 15 -X POST -H 'content-type: application/json' -d "{\"username\":\"$NU\",\"password\":\"SwitchTest123\"}" "$M/api/auth/register" -o /dev/null -w '%{http_code}')"
|
||
NID=$(Q "select id from users where username='$NU'")
|
||
echo " approve=$(curl -s -m 15 -X POST -b "$AC" "$M/api/admin/users/$NID/approve" -o /dev/null -w '%{http_code}')"
|
||
NC=$(curl -s -m 15 -X POST -H 'content-type: application/json' -d "{\"username\":\"$NU\",\"password\":\"SwitchTest123\"}" -D - "$M/api/auth/login" -o /dev/null | grep -i '^set-cookie' | head -1 | grep -oP 'sid=[^;]+')
|
||
echo " mkdir=$(curl -s -m 15 -X POST -b "$NC" -H 'content-type: application/json' -d '{"path":"proj"}' "$M/api/fs/mkdir" -o /dev/null -w '%{http_code}') ← 首次触达应把归属钉住"
|
||
echo " 钉住后归属: $(owner "$NU") ← 期望 w-106（与下面的 launch 必须同台）"
|
||
echo " upload=$(curl -s -m 20 -X POST -b "$NC" -H 'content-type: application/json' -d "{\"path\":\"proj\",\"name\":\"hello.txt\",\"data\":\"$(printf 'hi-from-switch' | base64 -w0)\"}" "$M/api/fs/upload" -o /dev/null -w '%{http_code}')"
|
||
echo " launch=$(curl -s -m 60 -X POST -b "$NC" -H 'content-type: application/json' -d '{"folder":"proj"}' "$M/api/dsh/launch" -o /dev/null -w '%{http_code}')"
|
||
sleep 4
|
||
echo " 归属: $(owner "$NU") ← 期望 w-106（粘性保持）"
|
||
echo " w-106=$(agent 19000 $T106) w-47=$(agent 19100 $T47)"
|
||
echo " --- 落盘取证 ---"
|
||
L47=$($SSH106 "ls /var/lib/dshs/users/$NID/ws/proj/hello.txt 2>/dev/null" 2>/dev/null || true)
|
||
echo " 106 盘: ${L47:-不存在}"
|
||
echo " 47 盘: $(ls /var/lib/dshs/users/$NID/ws/proj/hello.txt 2>/dev/null || echo 不存在（应不存在 ✓）)"
|
||
NP=$(curl -s -m 25 -L -b "$NC" -H "Host: $NU.$DOMAIN" "$M/" | head -c 300)
|
||
echo " 实例页(Host: $NU.$DOMAIN): $(isapp "$NP")"
|
||
|
||
echo
|
||
echo "########## 收尾 ##########"
|
||
curl -s -m 40 -X POST -b "$GC" "$M/api/dsh/stop" -o /dev/null -w " guest stop=%{http_code}\n"
|
||
curl -s -m 40 -X POST -b "$NC" "$M/api/dsh/stop" -o /dev/null -w " newuser stop=%{http_code}\n"
|
||
echo " stop 后 guest 归属（**不应再被清空**）: $(owner guest)"
|
||
Q "delete from sessions where user_agent='switch-verify'" >/dev/null
|
||
echo " 残留临时 session: $(Q "select count(*) from sessions where user_agent='switch-verify'")（应 0）"
|
||
echo " 新用户待清: $NU / $NID"
|