Files
heicode-mananger/heicode/main.go
T
chenchenandClaude Opus 4.8 2e37495133 feat(telemetry): retention purge + context field whitelist + size caps (#32)
Telemetry up-gating hardening (code portion of #32):

- Context field whitelist: telemetry `context` is filtered to a small set of
  non-content diagnostic keys (route/retryable/phase/exit_code/duration_ms/
  attempt) before persistence. Unknown keys — including potentially identifying
  ones (email, full file path, prompt, raw IP) — are dropped, so a client
  regression cannot land arbitrary JSON in the store. Empty/unparseable/no-allowed-key
  context is dropped to "".
- Per-field size cap: stack_top and context are truncated to 8KiB after
  redaction (backstop against unbounded blobs within batch limits).
- Retention: daily master-only task deletes telemetry rows older than
  HEICODE_TELEMETRY_RETENTION_DAYS (default 30; <=0 disables).
  HEICODE_TELEMETRY_RETENTION_INTERVAL_HOURS (default 24) sets cadence.
  model.DeleteTelemetryEventsBefore(cutoff) + controller.StartTelemetryRetentionTask()
  wired into main.go under IsMasterNode.
- GET /api/heicode/config telemetry block now surfaces retention_days for
  client/admin transparency.

Tests: whitelist drop/keep, size cap, redaction-within-allowed-key. go build/vet
clean; controller telemetry tests pass.

Affects: Manager only (telemetry ingest + retention). No billing/consume-log
change (telemetry still never bills). Privacy-doc disclosure + production
enable-checklist portions of #32 tracked in heicodeDocs sync (#34) / desktop
client API docs (#35).

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
2026-06-10 00:40:23 +08:00

395 lines
12 KiB
Go

package main
import (
"bytes"
"embed"
"fmt"
"log"
"net/http"
"os"
"strconv"
"strings"
"time"
"github.com/heicode/manager/common"
"github.com/heicode/manager/constant"
"github.com/heicode/manager/controller"
"github.com/heicode/manager/i18n"
"github.com/heicode/manager/logger"
"github.com/heicode/manager/middleware"
"github.com/heicode/manager/model"
"github.com/heicode/manager/oauth"
"github.com/heicode/manager/relay"
"github.com/heicode/manager/router"
"github.com/heicode/manager/service"
_ "github.com/heicode/manager/setting/performance_setting"
"github.com/heicode/manager/setting/ratio_setting"
"github.com/bytedance/gopkg/util/gopool"
"github.com/gin-contrib/sessions"
"github.com/gin-contrib/sessions/cookie"
"github.com/gin-gonic/gin"
"github.com/joho/godotenv"
_ "net/http/pprof"
)
//go:embed web/default/dist
var buildFS embed.FS
//go:embed web/default/dist/index.html
var indexPage []byte
//go:embed web/classic/dist
var classicBuildFS embed.FS
//go:embed web/classic/dist/index.html
var classicIndexPage []byte
func main() {
startTime := time.Now()
err := InitResources()
if err != nil {
common.FatalLog("failed to initialize resources: " + err.Error())
return
}
common.SysLog("New API " + common.Version + " started")
if os.Getenv("GIN_MODE") != "debug" {
gin.SetMode(gin.ReleaseMode)
}
if common.DebugEnabled {
common.SysLog("running in debug mode")
}
defer func() {
err := model.CloseDB()
if err != nil {
common.FatalLog("failed to close database: " + err.Error())
}
}()
if common.RedisEnabled {
// for compatibility with old versions
common.MemoryCacheEnabled = true
}
if common.MemoryCacheEnabled {
common.SysLog("memory cache enabled")
common.SysLog(fmt.Sprintf("sync frequency: %d seconds", common.SyncFrequency))
// Add panic recovery and retry for InitChannelCache
func() {
defer func() {
if r := recover(); r != nil {
common.SysLog(fmt.Sprintf("InitChannelCache panic: %v, retrying once", r))
// Retry once
_, _, fixErr := model.FixAbility()
if fixErr != nil {
common.FatalLog(fmt.Sprintf("InitChannelCache failed: %s", fixErr.Error()))
}
}
}()
model.InitChannelCache()
}()
go model.SyncChannelCache(common.SyncFrequency)
}
// 热更新配置
go model.SyncOptions(common.SyncFrequency)
// 数据看板
go model.UpdateQuotaData()
if os.Getenv("CHANNEL_UPDATE_FREQUENCY") != "" {
frequency, err := strconv.Atoi(os.Getenv("CHANNEL_UPDATE_FREQUENCY"))
if err != nil {
common.FatalLog("failed to parse CHANNEL_UPDATE_FREQUENCY: " + err.Error())
}
go controller.AutomaticallyUpdateChannels(frequency)
}
go controller.AutomaticallyTestChannels()
// Codex credential auto-refresh check every 10 minutes, refresh when expires within 1 day
service.StartCodexCredentialAutoRefreshTask()
// Subscription quota reset task (daily/weekly/monthly/custom)
service.StartSubscriptionQuotaResetTask()
// Wire task polling adaptor factory (breaks service -> relay import cycle)
service.GetTaskAdaptorFunc = func(platform constant.TaskPlatform) service.TaskPollingAdaptor {
a := relay.GetTaskAdaptor(platform)
if a == nil {
return nil
}
return a
}
// Channel upstream model update check task
controller.StartChannelUpstreamModelUpdateTask()
// Secret lifecycle: daily purge of vault secrets soft-deleted past the
// retention window (issue #4). Master-only so multiple nodes don't all purge.
if common.IsMasterNode {
controller.StartSecretPurgeTask()
// Telemetry retention: daily purge of client error-telemetry older than
// HEICODE_TELEMETRY_RETENTION_DAYS (#32). Master-only.
controller.StartTelemetryRetentionTask()
}
if common.IsMasterNode && constant.UpdateTask {
gopool.Go(func() {
controller.UpdateMidjourneyTaskBulk()
})
gopool.Go(func() {
controller.UpdateTaskBulk()
})
}
if os.Getenv("BATCH_UPDATE_ENABLED") == "true" {
common.BatchUpdateEnabled = true
common.SysLog("batch update enabled with interval " + strconv.Itoa(common.BatchUpdateInterval) + "s")
model.InitBatchUpdater()
}
if os.Getenv("ENABLE_PPROF") == "true" {
gopool.Go(func() {
log.Println(http.ListenAndServe("0.0.0.0:8005", nil))
})
go common.Monitor()
common.SysLog("pprof enabled")
}
err = common.StartPyroScope()
if err != nil {
common.SysError(fmt.Sprintf("start pyroscope error : %v", err))
}
// Initialize HTTP server
server := gin.New()
// Configure trusted proxy CIDRs so `c.ClientIP()` honours
// X-Forwarded-For / X-Real-IP set by the reverse proxy (Caddy /
// nginx / Cloudflare) that sits in front of the Manager. From
// Gin v1.7 the default is to NOT trust any header, which makes
// `c.ClientIP()` return the docker bridge peer (e.g. 10.2.3.4)
// — useless for audit trail, token IP allowlists, and rate
// limiting. Trusting the standard RFC1918 private ranges +
// loopback covers every realistic Manager deployment topology:
// - docker compose (host-network or bridge)
// - k8s ClusterIP service mesh
// - reverse proxy on same VM
// Public-IP proxies (e.g. Cloudflare edge IPs) are NOT in this
// list. If you front Manager directly with Cloudflare without
// a local reverse proxy, add the published CF ranges here.
// SetTrustedProxies returns an error only when the strings are
// not valid CIDRs — we panic because that's a config bug, not
// a runtime condition.
if err := server.SetTrustedProxies([]string{
"10.0.0.0/8", // RFC1918 + Azure VNet + docker bridge
"172.16.0.0/12", // docker bridge default
"192.168.0.0/16", // RFC1918 LAN
"127.0.0.1/32", // loopback IPv4
"::1/128", // loopback IPv6
"fd00::/8", // RFC4193 unique-local IPv6
}); err != nil {
common.FatalLog(fmt.Sprintf("SetTrustedProxies: %v", err))
}
// Cloudflare-aware ClientIP resolution. In production Manager runs
// behind Cloudflare's proxy mode, which strips the inbound
// X-Forwarded-For and replaces it with CF-Connecting-IP carrying
// the real client IP. Setting TrustedPlatform short-circuits Gin's
// default XFF parsing and reads CF-Connecting-IP as ground truth —
// without this, every request shows the docker bridge peer
// (10.2.3.x) regardless of how many CIDRs we add to TrustedProxies.
//
// Honoured globally: c.ClientIP() returns the real IP for all
// downstream consumers (token IP allowlist, rate-limit, device
// last-seen IP, gin access log, audit logs). When the header is
// absent (e.g. health checks from the VM itself, or direct
// non-CF hits) Gin falls back to TrustedProxies → XFF → RemoteAddr.
server.TrustedPlatform = gin.PlatformCloudflare
server.Use(gin.CustomRecovery(func(c *gin.Context, err any) {
common.SysLog(fmt.Sprintf("panic detected: %v", err))
c.JSON(http.StatusInternalServerError, gin.H{
"error": gin.H{
"message": fmt.Sprintf("Panic detected, error: %v. Please submit a issue here: https://github.com/Calcium-Ion/heicode", err),
"type": "new_api_panic",
},
})
}))
// This will cause SSE not to work!!!
//server.Use(gzip.Gzip(gzip.DefaultCompression))
server.Use(middleware.RequestId())
server.Use(middleware.PoweredBy())
server.Use(middleware.I18n())
middleware.SetUpLogger(server)
// Initialize session store
store := cookie.NewStore([]byte(common.SessionSecret))
store.Options(sessions.Options{
Path: "/",
MaxAge: 2592000, // 30 days
HttpOnly: true,
Secure: false,
SameSite: http.SameSiteStrictMode,
})
server.Use(sessions.Sessions("session", store))
InjectUmamiAnalytics()
InjectGoogleAnalytics()
// 设置路由
router.SetRouter(server, router.ThemeAssets{
DefaultBuildFS: buildFS,
DefaultIndexPage: indexPage,
ClassicBuildFS: classicBuildFS,
ClassicIndexPage: classicIndexPage,
})
var port = os.Getenv("PORT")
if port == "" {
port = strconv.Itoa(*common.Port)
}
// Log startup success message
common.LogStartupSuccess(startTime, port)
err = server.Run(":" + port)
if err != nil {
common.FatalLog("failed to start HTTP server: " + err.Error())
}
}
func InjectUmamiAnalytics() {
analyticsInjectBuilder := &strings.Builder{}
if os.Getenv("UMAMI_WEBSITE_ID") != "" {
umamiSiteID := os.Getenv("UMAMI_WEBSITE_ID")
umamiScriptURL := os.Getenv("UMAMI_SCRIPT_URL")
if umamiScriptURL == "" {
umamiScriptURL = "https://analytics.umami.is/script.js"
}
analyticsInjectBuilder.WriteString("<script defer src=\"")
analyticsInjectBuilder.WriteString(umamiScriptURL)
analyticsInjectBuilder.WriteString("\" data-website-id=\"")
analyticsInjectBuilder.WriteString(umamiSiteID)
analyticsInjectBuilder.WriteString("\"></script>")
}
analyticsInjectBuilder.WriteString("<!--Umami QuantumNous-->\n")
analyticsInject := []byte(analyticsInjectBuilder.String())
placeholder := []byte("<!--umami-->\n")
indexPage = bytes.ReplaceAll(indexPage, placeholder, analyticsInject)
classicIndexPage = bytes.ReplaceAll(classicIndexPage, placeholder, analyticsInject)
}
func InjectGoogleAnalytics() {
analyticsInjectBuilder := &strings.Builder{}
if os.Getenv("GOOGLE_ANALYTICS_ID") != "" {
gaID := os.Getenv("GOOGLE_ANALYTICS_ID")
// Google Analytics 4 (gtag.js)
analyticsInjectBuilder.WriteString("<script async src=\"https://www.googletagmanager.com/gtag/js?id=")
analyticsInjectBuilder.WriteString(gaID)
analyticsInjectBuilder.WriteString("\"></script>")
analyticsInjectBuilder.WriteString("<script>")
analyticsInjectBuilder.WriteString("window.dataLayer = window.dataLayer || [];")
analyticsInjectBuilder.WriteString("function gtag(){dataLayer.push(arguments);}")
analyticsInjectBuilder.WriteString("gtag('js', new Date());")
analyticsInjectBuilder.WriteString("gtag('config', '")
analyticsInjectBuilder.WriteString(gaID)
analyticsInjectBuilder.WriteString("');")
analyticsInjectBuilder.WriteString("</script>")
}
analyticsInjectBuilder.WriteString("<!--Google Analytics QuantumNous-->\n")
analyticsInject := []byte(analyticsInjectBuilder.String())
placeholder := []byte("<!--Google Analytics-->\n")
indexPage = bytes.ReplaceAll(indexPage, placeholder, analyticsInject)
classicIndexPage = bytes.ReplaceAll(classicIndexPage, placeholder, analyticsInject)
}
func InitResources() error {
// Initialize resources here if needed
// This is a placeholder function for future resource initialization
err := godotenv.Load(".env")
if err != nil {
if common.DebugEnabled {
common.SysLog("No .env file found, using default environment variables. If needed, please create a .env file and set the relevant variables.")
}
}
// 加载环境变量
common.InitEnv()
logger.SetupLogger()
// Initialize model settings
ratio_setting.InitRatioSettings()
service.InitHttpClient()
service.InitTokenEncoders()
// Initialize SQL Database
err = model.InitDB()
if err != nil {
common.FatalLog("failed to initialize database: " + err.Error())
return err
}
model.CheckSetup()
// Initialize options, should after model.InitDB()
model.InitOptionMap()
// 清理旧的磁盘缓存文件
common.CleanupOldCacheFiles()
// 初始化模型
model.GetPricing()
// Initialize SQL Database
err = model.InitLogDB()
if err != nil {
return err
}
// Initialize Redis
err = common.InitRedisClient()
if err != nil {
return err
}
// Bootstrap the long-term X25519 keypair used by V2 device-bound
// clients to encrypt request bodies. Generates on first launch,
// reuses persisted record otherwise. Required before any /v1/*
// traffic that carries Content-Encoding: heicode-aead-v1 can be
// decrypted.
if err := service.EnsureServerECDHKey(); err != nil {
common.SysError("ECDH server key bootstrap failed: " + err.Error())
return err
}
// 启动系统监控
common.StartSystemMonitor()
// Initialize i18n
err = i18n.Init()
if err != nil {
common.SysError("failed to initialize i18n: " + err.Error())
// Don't return error, i18n is not critical
} else {
common.SysLog("i18n initialized with languages: " + strings.Join(i18n.SupportedLanguages(), ", "))
}
// Register user language loader for lazy loading
i18n.SetUserLangLoader(model.GetUserLanguage)
// Load custom OAuth providers from database
err = oauth.LoadCustomProviders()
if err != nil {
common.SysError("failed to load custom OAuth providers: " + err.Error())
// Don't return error, custom OAuth is not critical
}
return nil
}