memory-eternal 0.10.8 → 0.10.10

package/lib/config-sync.js

@@ -45,7 +45,7 @@ export function readPendingConfig(env = process.env) {
  * @param {string} text 完整内容
  * @returns {string} 目标文件路径
  */
-export function writeFileAtomicSync(file, text, { retries = 8, delayMs = 20 } = {}) {
+export function writeFileAtomicSync(file, text, { retries = 12, delayMs = 25, onDegraded } = {}) {
   const tmp = file + '.' + process.pid + '.tmp'
   fs.writeFileSync(tmp, text, 'utf8')
   const retriable = (code) => code === 'EPERM' || code === 'EBUSY' || code === 'EACCES'
@@ -60,6 +60,10 @@ export function writeFileAtomicSync(file, text, { retries = 8, delayMs = 20 } =
       if (i < retries) { sleepSync(delayMs); continue }
       // 退避仍失败（极端争用）：退回直接覆盖写。宁可短暂退化到非原子，也不能把用户的
       // 配置改动整份丢掉 —— 这是这个函数存在的第一原则。
+      //
+      // 这是**可观测的退化**：并发读方在这个窗口里可能读到半截。tests/config-sync 的并发用例
+      // 靠 onDegraded 区分「原子路径失效」与「OS 拒绝了替换」—— 前者仍然必须红。
+      if (typeof onDegraded === 'function') { try { onDegraded(file) } catch { /* 回调异常不影响写入 */ } }
       fs.writeFileSync(file, text, 'utf8')
       try { fs.unlinkSync(tmp) } catch { /* 已清理 */ }
       return file

package/lib/watchdog.js

@@ -448,12 +448,19 @@ export async function waitForServedVersion(port, expect, { timeoutMs = 15000, in
  * @returns {Promise<{port:number, served:string, occupant:{pid:number,command:string}|null,
  *   occupantIsOurs:boolean, watchdog:object|null, watchdogAlive:boolean, residentStartedAt:number}>}
  */
-export async function inspectResident({ env = process.env, port, probeVersion, findListener } = {}) {
+export async function inspectResident({ env = process.env, port, probeVersion, findListener, retries = 3, retryDelayMs = 500, sleepFn } = {}) {
   const p = Number(port) || 0
   const entry = readWatchdogLock(env).watchdogs.find((w) => Number(w.port) === p && isPidAlive(w.pid)) || null
   let served = ''
   if (p) {
-    try { served = String((await (probeVersion || probeServedVersion)(p)) || '') } catch { served = '' }
+    // issue #29 建议 3：一次探针打空（实例正忙 / GC / 冷启动）**不能**当成「旧版不自报版本」——
+    // 那个结论会直接触发漂移替换，于是每次启动都多起一次常驻实例。退避重试后再下结论。
+    const wait = sleepFn || ((ms) => new Promise((r) => setTimeout(r, ms)))
+    for (let i = 0; ; i++) {
+      try { served = String((await (probeVersion || probeServedVersion)(p)) || '') } catch { served = '' }
+      if (served || i >= Math.max(0, Number(retries) || 0)) break
+      await wait(Math.max(0, Number(retryDelayMs) || 0))
+    }
   }
   let occupant = null
   // 只有「探不到版本」时才需要问端口占用者：能自报版本就已经证明是我们的服务了

package/lib/web.js

@@ -246,13 +246,42 @@ export async function probeWebServer(port = DEFAULT_WEB_PORT, timeoutMs = 800) {
   }
 }
 
+/**
+ * 宽松判定：这个端口上是不是**自家的 web**（旧版可能不自报 version / vaultDir，严格探针会判否）。
+ *
+ * 为什么需要（issue #29）：升级期端口被**自家旧包**的 detached 实例占着时，严格探针认不出它，
+ * `ensureWebServer()` 就会向后漂到一个空端口 → 一段窗口里同时存在两个实例、两个端口，
+ * 而宿主记录的 `web-info` 指向的可能是随后被自愈收掉的那个。多探几个自家路由即可识别。
+ */
+export async function looksLikeOurWeb(port = DEFAULT_WEB_PORT, timeoutMs = 600) {
+  for (const suffix of ['/overview', '/web-info', '/budget']) {
+    try {
+      const res = await fetch(`http://127.0.0.1:${port}${API_PREFIX}${suffix}`, { signal: AbortSignal.timeout(timeoutMs) })
+      if (!res.ok) continue
+      const data = await res.json().catch(() => null)
+      if (data && typeof data === 'object' && data.ok === true) return true
+    } catch { /* 换下一个探针 */ }
+  }
+  return false
+}
+
+/** 等端口上的自家实例变成「可严格识别」（自愈替换完成）—— 期间它可能被清掉再拉起。 */
+async function waitForOurWeb(port, timeoutMs = 6000, intervalMs = 500) {
+  const deadline = Date.now() + Math.max(0, Number(timeoutMs) || 0)
+  for (;;) {
+    if (await probeWebServer(port)) return true
+    if (Date.now() >= deadline) return false
+    await new Promise((r) => setTimeout(r, Math.max(100, Number(intervalMs) || 500)))
+  }
+}
+
 /**
  * 确保 web server 存活（默认开启的实现核心）。
  * 已活 → 直接返回 URL；未活 → detached spawn 独立进程并轮询探活。
  * 端口被非本项目进程占用时向后漂移（+1..+10）。
  * @returns {Promise<{url:string, port:number, spawned:boolean}>}
  */
-export async function ensureWebServer({ port = DEFAULT_WEB_PORT, vaultRoot = defaultVaultDir(), spawnEnv = {}, totalTimeoutMs = 15000 } = {}) {
+export async function ensureWebServer({ port = DEFAULT_WEB_PORT, vaultRoot = defaultVaultDir(), spawnEnv = {}, totalTimeoutMs = 15000, takeoverWaitMs = 6000 } = {}) {
   // 探测可用端口：先找「已是我们的服务」的端口，再找空闲端口
   let target = port
   let found = null
@@ -260,8 +289,16 @@ export async function ensureWebServer({ port = DEFAULT_WEB_PORT, vaultRoot = def
     const alive = await probeWebServer(p)
     if (alive) { found = { url: `http://127.0.0.1:${p}`, port: p, spawned: false }; break }
     // 该端口不是我们的服务：若连不上（空闲）就用它；连得上但不是我们的（被占用）则漂移
-    const isFree = await probeWebServer(p, 300) === null && await isPortConnectable(p) === false
-    if (isFree) { target = p; break }
+    if (await isPortConnectable(p) === false) { target = p; break }
+    // issue #29：占用者「像自家的 web、但严格探针认不出」（升级期旧包的 detached 实例）时，
+    // **不要立刻漂到 port+1** —— 自愈（restartResident）马上会把配置端口上的实例换成新版，
+    // 这时另起一个端口只会多出一个实例、并让 web-info 指向随后被收掉的那个（客户端 iframe
+    // 于是「拒绝连接」）。先等它接管；等不到再漂，并留下一条能查的日志。
+    if (await looksLikeOurWeb(p)) {
+      if (await waitForOurWeb(p, takeoverWaitMs)) { found = { url: `http://127.0.0.1:${p}`, port: p, spawned: false }; break }
+      if (await isPortConnectable(p) === false) { target = p; break }
+      process.stderr.write(`[memory-eternal] 端口 ${p} 被自家旧实例占着且 ${takeoverWaitMs}ms 内没让位 → 向后找空闲端口（web-info 会实时收敛，issue #29）\n`)
+    }
     target = p + 1
   }
   if (found) return found

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "memory-eternal",
-  "version": "0.10.8",
+  "version": "0.10.10",
   "description": "记忆核心（Memory Eternal）：自研的 DeepSeek Harness 记忆插件，不移植任何既有记忆框架——对话结束后自动沉淀知识卡到本地 Markdown Vault（自研去重、自研 CJK 检索、可 git 管理），设置页提供图形化知识库（统计 / 搜索 / 知识图谱 + 侧边栏一键弹窗），Agent 通过 memory_recall 工具按需召回历史上下文。零人工干预。",
   "license": "MIT",
   "repository": {
@@ -64,7 +64,7 @@
     "prepublishOnly": "node build.mjs",
     "release": "node scripts/release.mjs",
     "release:dry": "node scripts/release.mjs --dry-run",
-    "test": "node build.mjs && node tests/vault.test.mjs && node tests/capture.test.mjs && node tests/capture-route.test.mjs && node tests/stall.test.mjs && node tests/vault-path.test.mjs && node tests/vault-resolve.test.mjs && node tests/minhash.test.mjs && node tests/settings-save.test.mjs && node tests/config-coverage.test.mjs && node tests/config-sync.test.mjs && node tests/config-merge.test.mjs && node tests/sse.test.mjs && node tests/api.test.mjs && node tests/import-roundtrip.test.mjs && node tests/graph-perf.test.mjs && node tests/settings-compat.test.mjs && node tests/node-bin.test.mjs && node tests/watchdog-lock.test.mjs && node tests/markdown.test.mjs && node tests/feedback.test.mjs && node tests/graph-lod.test.mjs && node tests/ssr.test.mjs && node tests/web-assets.test.mjs && node tests/manifests.test.mjs && node tests/two-store.test.mjs && node tests/noise-gate.test.mjs && node tests/e2e-recall.test.mjs && node tests/issue-fixes.test.mjs && node tests/issue-fixes-2.test.mjs && node tests/issue-fixes-3.test.mjs && node tests/watchdog-orphans.test.mjs && node tests/watchdog-restart.test.mjs && node tests/readme-i18n.test.mjs && node tests/issue-fixes-4.test.mjs && node tests/host-drain.test.mjs && node tests/e2e-host-drain.mjs"
+    "test": "node build.mjs && node tests/vault.test.mjs && node tests/capture.test.mjs && node tests/capture-route.test.mjs && node tests/stall.test.mjs && node tests/vault-path.test.mjs && node tests/vault-resolve.test.mjs && node tests/minhash.test.mjs && node tests/settings-save.test.mjs && node tests/config-coverage.test.mjs && node tests/config-sync.test.mjs && node tests/config-merge.test.mjs && node tests/sse.test.mjs && node tests/api.test.mjs && node tests/import-roundtrip.test.mjs && node tests/graph-perf.test.mjs && node tests/settings-compat.test.mjs && node tests/node-bin.test.mjs && node tests/watchdog-lock.test.mjs && node tests/markdown.test.mjs && node tests/feedback.test.mjs && node tests/graph-lod.test.mjs && node tests/ssr.test.mjs && node tests/web-assets.test.mjs && node tests/manifests.test.mjs && node tests/two-store.test.mjs && node tests/noise-gate.test.mjs && node tests/e2e-recall.test.mjs && node tests/issue-fixes.test.mjs && node tests/issue-fixes-2.test.mjs && node tests/issue-fixes-3.test.mjs && node tests/watchdog-orphans.test.mjs && node tests/watchdog-restart.test.mjs && node tests/readme-i18n.test.mjs && node tests/issue-fixes-4.test.mjs && node tests/host-drain.test.mjs && node tests/e2e-host-drain.mjs && node tests/issue-fixes-5.test.mjs && node tests/issue-fixes-6.test.mjs"
   },
   "dependencies": {
     "@deepseek-ai/schemastery": "^3.18.1"