mirror of
https://gitcode.com/JianFeeeee/HomeAgent.git
synced 2026-09-21 09:28:14 +00:00
根因:子进程插件被 kill 后,内核只发了一个无人订阅的事件, 工具/stage handler/IO 通道全留在注册表里指向死进程, 模型继续调用只吃 ErrProcessExited,没有任何路径把插件拉回来。 ## 四层修复 ### 1. 专职 waitLoop(进程收割) - 每个子进程配一根 waitLoop goroutine,是 cmd.Wait() 的唯一调用点 - 不再依赖 stdout EOF 判定死亡(孙子进程继承 stdout 时 EOF 永不到来) - 手工 os.Pipe 替代 cmd.StdinPipe/StdoutPipe,避免 waitLoop 与 os/exec 的内部关闭竞争 - host.go: Host.Supervisor(),Host.Close() 先 StopAll 再拆段 ### 2. 集中台账 Supervisor - proc/supervisor.go: 插件 Spawn 握手成功即 track,进程退出即 untrack - StopAll: 并发发 plugin.stop 走优雅路径,到期仍在的一律 Kill - 关停后才完成握手的进程被立即结束,不会活过内核 - 消除「孤儿进程持共享段映射 → SIGBUS」的隐患 ### 3. 注册面摘除(detachPlugin) - 新增 StageHost.UnregisterPluginStages:摘除指定插件的全部 stage handler - 新增 Registry.pluginChannels 台账:记录每个插件注册的 IO 通道 - 三条路径统一走 detachPlugin:Disable / ReloadOne / RemovePlugin - StopAndUnload 漏了 IO 通道也一并补上 ### 4. 自动重启 - onProcCrash 从「只发事件」改为「摘注册面 → 从注册表移除 → 异步排重启」 - scheduleProcRestart: 窗口 5 分钟内最多 3 次,线性退避 1s/2s/3s - 超限停手留日志;重启前复核是否已被 Disable 或被其他路径加载 - 崩溃计数窗口过期自动归零 ### 5. 主动停止 vs 崩溃的区分 - proc.Plugin 新增 stopping 标志:Stop()/Close() 里 Set(true) - handleExit 读 stopping 标志,主动停止不上报 onCrash - 防止重载/禁用/卸载被误判为崩溃触发多余重启 ### 6. Linux Pdeathsig 兜底 - procattr_linux.go: SysProcAttr.Pdeathsig = SIGKILL - 兜 homed 自身被 SIGKILL/OOM 时子进程变孤儿的场景 - macOS/Windows 无等价物,空实现 ### 7. pluginmgr 升级 - PluginManager 接口新增 PluginRuntime / ListPluginRuntimes - plugin_list 输出运行态:loaded / alive / pid / crash_count / channel - 新增 plugin_status: 全量运行期快照 + dead/unhealthy 汇总 - 新增 plugin_restart: 无条件重启单个插件(plgreload 不动未改二进制的插件) ### 测试 - process_test.go: 3 例(grandchild stdout 感知 / Supervisor track-untrack / StopAll 无孤儿) - crash_recovery_test.go: 8 例(detach 三项齐全 / 通道重注册 / 崩溃不阻塞 / 退避阈值 / 窗口过期 / 关停中跳过 / PluginRuntime 通道识别) - stages_plugin_test.go: 4 例(stage 按插件摘除 / 空 stage 清理 / 空名 no-op / 工具+stage 双摘后可重新注册同名)
209 lines
6.1 KiB
Go
209 lines
6.1 KiB
Go
package webui
|
||
|
||
import (
|
||
"net/http"
|
||
"net/http/httptest"
|
||
"testing"
|
||
|
||
sdk "gitcode.com/JianFeeeee/HomeAgent/internal/sdk"
|
||
)
|
||
|
||
// mockPluginMgr 是 sdk.PluginManager 的最小实现,用于 handler 路由层测试。
|
||
type mockPluginMgr struct {
|
||
builtins map[string]bool
|
||
disabled []sdk.DisabledPluginInfo
|
||
isDisabled map[string]bool
|
||
runtimes map[string]sdk.PluginRuntimeInfo
|
||
|
||
removed []string
|
||
reloadN int
|
||
removeErr error
|
||
}
|
||
|
||
func (m *mockPluginMgr) ListLoadedPlugins() []string { return nil }
|
||
func (m *mockPluginMgr) ListDisabledPlugins() []sdk.DisabledPluginInfo {
|
||
return m.disabled
|
||
}
|
||
func (m *mockPluginMgr) IsPluginDisabled(name string) bool {
|
||
return m.isDisabled[name]
|
||
}
|
||
func (m *mockPluginMgr) IsBuiltinPlugin(name string) bool {
|
||
return m.builtins[name]
|
||
}
|
||
func (m *mockPluginMgr) DisablePlugin(name, by string) error {
|
||
if m.isDisabled == nil {
|
||
m.isDisabled = map[string]bool{}
|
||
}
|
||
m.isDisabled[name] = true
|
||
return nil
|
||
}
|
||
func (m *mockPluginMgr) EnablePlugin(name string) error {
|
||
delete(m.isDisabled, name)
|
||
return nil
|
||
}
|
||
func (m *mockPluginMgr) RemovePlugin(name string) error {
|
||
if m.removeErr != nil {
|
||
return m.removeErr
|
||
}
|
||
m.removed = append(m.removed, name)
|
||
return nil
|
||
}
|
||
func (m *mockPluginMgr) ReloadPlugins() (string, error) {
|
||
m.reloadN++
|
||
return "reloaded", nil
|
||
}
|
||
func (m *mockPluginMgr) ReloadOne(name string) error { return nil }
|
||
func (m *mockPluginMgr) StopAndUnload(name string) error { return nil }
|
||
func (m *mockPluginMgr) PluginMetas() map[string]sdk.PluginMeta {
|
||
return nil
|
||
}
|
||
func (m *mockPluginMgr) PluginDir() string { return "" }
|
||
func (m *mockPluginMgr) PluginRuntime(name string) (sdk.PluginRuntimeInfo, bool) {
|
||
info, ok := m.runtimes[name]
|
||
return info, ok
|
||
}
|
||
func (m *mockPluginMgr) ListPluginRuntimes() []sdk.PluginRuntimeInfo {
|
||
out := make([]sdk.PluginRuntimeInfo, 0, len(m.runtimes))
|
||
for _, v := range m.runtimes {
|
||
out = append(out, v)
|
||
}
|
||
return out
|
||
}
|
||
|
||
// newHandlerWithMock 构造带 mock PluginManager 的 Handler(绕过 SDK 组装)。
|
||
func newHandlerWithMock(m *mockPluginMgr) *Handler {
|
||
h := NewHandler(nil)
|
||
h.pluginMgr = m
|
||
return h
|
||
}
|
||
|
||
func TestHandlePluginByID_DisabledList(t *testing.T) {
|
||
m := &mockPluginMgr{
|
||
disabled: []sdk.DisabledPluginInfo{{Name: "foo"}},
|
||
}
|
||
h := newHandlerWithMock(m)
|
||
|
||
req := httptest.NewRequest(http.MethodGet, "/api/v1/plugins/disabled", nil)
|
||
w := httptest.NewRecorder()
|
||
h.handlePluginByID(w, req)
|
||
|
||
if w.Code != http.StatusOK {
|
||
t.Fatalf("expected 200, got %d", w.Code)
|
||
}
|
||
if body := w.Body.String(); len(body) == 0 || !contains(body, "foo") {
|
||
t.Fatalf("expected disabled list with foo, got %s", body)
|
||
}
|
||
}
|
||
|
||
func contains(s, sub string) bool {
|
||
return len(s) >= len(sub) && (func() bool {
|
||
for i := 0; i+len(sub) <= len(s); i++ {
|
||
if s[i:i+len(sub)] == sub {
|
||
return true
|
||
}
|
||
}
|
||
return false
|
||
})()
|
||
}
|
||
|
||
func TestHandlePluginByID_ReloadPost(t *testing.T) {
|
||
m := &mockPluginMgr{}
|
||
h := newHandlerWithMock(m)
|
||
|
||
req := httptest.NewRequest(http.MethodPost, "/api/v1/plugins/reload", nil)
|
||
w := httptest.NewRecorder()
|
||
h.handlePluginByID(w, req)
|
||
|
||
if w.Code != http.StatusOK {
|
||
t.Fatalf("expected 200, got %d: %s", w.Code, w.Body.String())
|
||
}
|
||
if m.reloadN != 1 {
|
||
t.Fatalf("expected ReloadPlugins called once, got %d", m.reloadN)
|
||
}
|
||
}
|
||
|
||
func TestHandlePluginByID_ReloadDeleteRejected(t *testing.T) {
|
||
// DELETE /plugins/reload 不允许把保留字当插件名反代成卸载
|
||
m := &mockPluginMgr{}
|
||
h := newHandlerWithMock(m)
|
||
|
||
req := httptest.NewRequest(http.MethodDelete, "/api/v1/plugins/reload", nil)
|
||
w := httptest.NewRecorder()
|
||
h.handlePluginByID(w, req)
|
||
|
||
if w.Code != http.StatusMethodNotAllowed {
|
||
t.Fatalf("expected 405, got %d", w.Code)
|
||
}
|
||
if len(m.removed) != 0 {
|
||
t.Fatalf("expected no plugin removed, got %v", m.removed)
|
||
}
|
||
}
|
||
|
||
func TestHandlePluginByID_PathTraversalRejected(t *testing.T) {
|
||
cases := []string{"../evil", "a/b", "a..b-ok-but-dots-only-check", "..%2Fetc"}
|
||
for _, name := range cases {
|
||
req := httptest.NewRequest(http.MethodDelete, "/api/v1/plugins/"+name, nil)
|
||
w := httptest.NewRecorder()
|
||
h := newHandlerWithMock(&mockPluginMgr{})
|
||
h.handlePluginByID(w, req)
|
||
// 含路径分隔符或以 .. 开头的名称必须被拒绝(400/405),绝不能反代到 pluginmgr
|
||
if w.Code != http.StatusBadRequest && w.Code != http.StatusMethodNotAllowed && w.Code != http.StatusNotFound {
|
||
t.Errorf("name %q: expected 4xx rejection, got %d", name, w.Code)
|
||
}
|
||
}
|
||
}
|
||
|
||
func TestHandlePluginByID_InvalidNamesRejected(t *testing.T) {
|
||
cases := []string{"has%20space", "has%3Acolon", "back%5Cslash"}
|
||
for _, name := range cases {
|
||
req := httptest.NewRequest(http.MethodGet, "/api/v1/plugins/"+name, nil)
|
||
w := httptest.NewRecorder()
|
||
h := newHandlerWithMock(&mockPluginMgr{})
|
||
h.handlePluginByID(w, req)
|
||
if w.Code != http.StatusBadRequest {
|
||
t.Errorf("name %q: expected 400, got %d", name, w.Code)
|
||
}
|
||
}
|
||
}
|
||
|
||
func TestHandlePluginByID_ValidNamePassesValidation(t *testing.T) {
|
||
// 合法插件名(含点/横线/下划线)不应被名称校验拦截;
|
||
// 这里 pluginmgr 未运行会得到 502 Bad Gateway,但绝不应该是 400。
|
||
m := &mockPluginMgr{}
|
||
h := newHandlerWithMock(m)
|
||
|
||
for _, name := range []string{"my-plugin", "plugin_v2", "weather.so"} {
|
||
req := httptest.NewRequest(http.MethodGet, "/api/v1/plugins/"+name, nil)
|
||
w := httptest.NewRecorder()
|
||
h.handlePluginByID(w, req)
|
||
if w.Code == http.StatusBadRequest {
|
||
t.Errorf("valid name %q should pass validation, got 400", name)
|
||
}
|
||
}
|
||
}
|
||
|
||
func TestHandlePluginByID_DisableEnable(t *testing.T) {
|
||
m := &mockPluginMgr{builtins: map[string]bool{"webui": true}}
|
||
h := newHandlerWithMock(m)
|
||
|
||
req := httptest.NewRequest(http.MethodPost, "/api/v1/plugins/webui/disable", nil)
|
||
w := httptest.NewRecorder()
|
||
h.handlePluginByID(w, req)
|
||
if w.Code != http.StatusOK {
|
||
t.Fatalf("disable: expected 200, got %d: %s", w.Code, w.Body.String())
|
||
}
|
||
if !m.isDisabled["webui"] {
|
||
t.Fatal("webui should be disabled in mock")
|
||
}
|
||
|
||
req = httptest.NewRequest(http.MethodPost, "/api/v1/plugins/webui/enable", nil)
|
||
w = httptest.NewRecorder()
|
||
h.handlePluginByID(w, req)
|
||
if w.Code != http.StatusOK {
|
||
t.Fatalf("enable: expected 200, got %d: %s", w.Code, w.Body.String())
|
||
}
|
||
if m.isDisabled["webui"] {
|
||
t.Fatal("webui should be re-enabled")
|
||
}
|
||
}
|