keeporg/cmd/keeporgd/main.go

// Command keeporgd is the KeepOrg synchronization server.
//
// It stores KDBX files as opaque, versioned blobs and arbitrates concurrent
// writes. It never sees a master password and cannot read a database.
package main

import (
	"context"
	"crypto/tls"
	"encoding/json"
	"errors"
	"flag"
	"fmt"
	"log"
	"net"
	"net/http"
	"os"
	"os/signal"
	"path/filepath"
	"strings"
	"syscall"
	"time"

	"cc.tiredsysadmin.keeporg/internal/kdbx"
	"cc.tiredsysadmin.keeporg/internal/secret"
	"cc.tiredsysadmin.keeporg/internal/server"
	"cc.tiredsysadmin.keeporg/internal/store"
)

// Settings a deployment usually knows only at deploy time can come from the
// environment as well. Each sits between the configuration file and its flag.
const (
	// externalURLEnv sets external_url, overridden by -url.
	externalURLEnv = "KEEPORG_EXTERNAL_URL"
	// databaseEnv sets database, overridden by -database. A PostgreSQL URL
	// carries a password, which is why it belongs in a secret, not a file.
	databaseEnv = "KEEPORG_DATABASE_URL"
	// proxiesEnv sets trusted_proxies, comma-separated.
	proxiesEnv = "KEEPORG_TRUSTED_PROXIES"
)

// Version is stamped at build time from versions.mk.
var Version = "dev"

// fileConfig is the on-disk shape of the configuration. It keeps durations as
// plain numbers so the file stays easy to edit by hand.
type fileConfig struct {
	Addr            string            `json:"addr"`
	DataDir         string            `json:"data_dir"`
	ExternalURL     string            `json:"external_url"`
	SessionTTLHours int               `json:"session_ttl_hours"`
	MaxDBSizeMB     int64             `json:"max_db_size_mb"`
	TLSCert         string            `json:"tls_cert"`
	TLSKey          string            `json:"tls_key"`
	Database        string            `json:"database"`
	TrustedProxies  []string          `json:"trusted_proxies"`
	MasterKeyFile   string            `json:"master_key_file"`
	VaultKDF        vaultKDF          `json:"vault_kdf"`
	OIDC            server.OIDCConfig `json:"oidc"`
}

// vaultKDF is the work factor for vaults this server creates, in the plain
// units an operator wants to edit.
type vaultKDF struct {
	MemoryMB   uint32 `json:"memory_mb"`
	Iterations uint32 `json:"iterations"`
	Lanes      uint8  `json:"lanes"`
}

func (f fileConfig) toConfig() server.Config {
	c := server.Config{
		Addr: f.Addr, DataDir: f.DataDir, ExternalURL: f.ExternalURL,
		TLSCert: f.TLSCert, TLSKey: f.TLSKey, OIDC: f.OIDC,
		Database: f.Database, MasterKeyFile: f.MasterKeyFile, TrustedProxies: f.TrustedProxies,
	}
	if f.VaultKDF.MemoryMB > 0 {
		c.VaultKDF = kdbx.KDFParams{
			MemoryKiB:  f.VaultKDF.MemoryMB * 1024,
			Iterations: f.VaultKDF.Iterations,
			Lanes:      f.VaultKDF.Lanes,
		}
	}
	if f.SessionTTLHours > 0 {
		c.SessionTTL = time.Duration(f.SessionTTLHours) * time.Hour
	}
	if f.MaxDBSizeMB > 0 {
		c.MaxDBSize = f.MaxDBSizeMB << 20
	}
	return c
}

func main() {
	log.SetFlags(log.LstdFlags | log.LUTC)
	lg := log.New(os.Stderr, "keeporgd: ", log.LstdFlags|log.LUTC)

	var (
		configPath = flag.String("config", "", "path to a JSON configuration file")
		addr       = flag.String("addr", "", "listen address (overrides the config file)")
		dataDir    = flag.String("data", "", "data directory (overrides the config file)")
		extURL     = flag.String("url", "", "externally reachable base URL (overrides $"+externalURLEnv+" and the config file)")
		database   = flag.String("database", "", "postgres://… or sqlite:<path>; default sqlite:<data>/keeporg.db "+
			"(overrides $"+databaseEnv+" and the config file)")
		masterKey = flag.String("master-key-file", "",
			"file holding the master key that wraps the vault keys, created if "+
				"absent; $"+secret.MasterKeyEnv+" and $"+secret.MasterKeyFileEnv+
				" take precedence over it and are never created")
		showVer    = flag.Bool("version", false, "print the version and exit")
		healthOnly = flag.Bool("healthcheck", false,
			"probe a running server on this address and exit 0 if it is healthy")
		writeCfg = flag.String("write-config", "",
			"write a commented example config to this path and exit")
		resetAdmin = flag.Bool("reset-admin", false,
			"reset the built-in administrator's password, print a new one and "+
				"exit; a running server picks it up within seconds")
	)
	flag.Parse()

	if *showVer {
		fmt.Println("keeporgd", Version)
		return
	}
	if *writeCfg != "" {
		if err := writeExampleConfig(*writeCfg); err != nil {
			lg.Fatal(err)
		}
		fmt.Println("wrote", *writeCfg)
		return
	}

	var fc fileConfig
	if *configPath != "" {
		b, err := os.ReadFile(*configPath)
		if err != nil {
			lg.Fatalf("read config: %v", err)
		}
		if err := json.Unmarshal(b, &fc); err != nil {
			lg.Fatalf("parse config: %v", err)
		}
	}
	cfg := fc.toConfig()
	// The public URL is the one setting a deployment usually knows only at
	// deploy time, so it can come from the environment as well. A flag
	// still wins.
	if v := strings.TrimSpace(os.Getenv(externalURLEnv)); v != "" {
		cfg.ExternalURL = v
	}
	if v := strings.TrimSpace(os.Getenv(databaseEnv)); v != "" {
		cfg.Database = v
	}
	if v := strings.TrimSpace(os.Getenv(proxiesEnv)); v != "" {
		cfg.TrustedProxies = strings.Split(v, ",")
	}
	if *database != "" {
		cfg.Database = *database
	}
	if *addr != "" {
		cfg.Addr = *addr
	}
	if *dataDir != "" {
		cfg.DataDir = *dataDir
	}
	if *extURL != "" {
		cfg.ExternalURL = *extURL
	}
	if *masterKey != "" {
		cfg.MasterKeyFile = *masterKey
	}

	if *healthOnly {
		// The container image has no shell, so the binary probes itself.
		if err := healthcheck(cfg); err != nil {
			fmt.Fprintln(os.Stderr, "unhealthy:", err)
			os.Exit(1)
		}
		return
	}

	if *resetAdmin {
		if err := resetAdministrator(cfg, lg); err != nil {
			lg.Fatalf("reset-admin: %v", err)
		}
		return
	}

	srv, err := server.New(cfg, lg)
	if err != nil {
		lg.Fatalf("start: %v", err)
	}
	defer srv.Close()
	if err := bootstrap(srv, lg); err != nil {
		lg.Fatalf("bootstrap: %v", err)
	}

	done := make(chan struct{})
	srv.StartHousekeeping(done)

	hs := &http.Server{
		Addr:              srv.Config().Addr,
		Handler:           srv.Handler(),
		ReadHeaderTimeout: 15 * time.Second,
		// No write timeout: the event stream is long lived by design.
		IdleTimeout: 120 * time.Second,
	}

	go func() {
		eff := srv.Config()
		lg.Printf("version %s listening on %s (external %s, database %s)",
			Version, eff.Addr, eff.ExternalURL, srv.DatabaseWhere())
		var err error
		if eff.TLSCert != "" && eff.TLSKey != "" {
			err = hs.ListenAndServeTLS(eff.TLSCert, eff.TLSKey)
		} else {
			err = hs.ListenAndServe()
		}
		if err != nil && !errors.Is(err, http.ErrServerClosed) {
			lg.Fatalf("listen: %v", err)
		}
	}()

	sig := make(chan os.Signal, 1)
	signal.Notify(sig, syscall.SIGINT, syscall.SIGTERM)
	<-sig
	close(done)
	ctx, cancel := context.WithTimeout(context.Background(), 10*time.Second)
	defer cancel()
	if err := hs.Shutdown(ctx); err != nil {
		lg.Printf("shutdown: %v", err)
	}
	lg.Print("stopped")
}

// bootstrap creates the first administrator and a root group on an empty
// server, so a fresh deployment is immediately usable. It is idempotent.
func bootstrap(srv *server.Server, lg *log.Logger) error {
	password := os.Getenv("KEEPORG_ADMIN_PASSWORD")
	generated := false
	if password == "" {
		password = secret.Token()[:20]
		generated = true
	}
	var created bool
	err := srv.Store().Update(func(st *store.State) error {
		if len(st.Users) > 0 {
			return nil
		}
		hash, err := secret.Hash(password)
		if err != nil {
			return err
		}
		u := &store.User{ID: secret.ID(), Username: "admin", PassHash: hash,
			ServerAdmin: true, CreatedAt: time.Now()}
		st.Users[u.ID] = u
		g := &store.Group{ID: secret.ID(), Name: "everyone", CreatedAt: time.Now()}
		st.Groups[g.ID] = g
		st.Memberships = append(st.Memberships,
			store.Membership{UserID: u.ID, GroupID: g.ID, Role: store.RoleAdmin})
		created = true
		return nil
	})
	if err != nil || !created {
		return err
	}
	if generated {
		lg.Printf("created the first administrator: username %q password %q "+
			"(shown once, change it or enrol TOTP now)", "admin", password)
	} else {
		lg.Print(`created the first administrator: username "admin" with the password from KEEPORG_ADMIN_PASSWORD`)
	}
	return nil
}

// resetAdministrator gives the server's administrator a new password, for
// when the one printed at first start was not kept. It writes to the
// database like any server process would, so a running server picks the
// change up within seconds and need not be stopped.
//
// Any second factor on that account is cleared too. Someone who has lost
// the password has often lost the authenticator with it, and a recovery
// that leaves you locked out is not one.
func resetAdministrator(cfg server.Config, lg *log.Logger) error {
	cfg.Defaults()
	kr, err := secret.LoadKeyring(cfg.MasterKeyFile)
	if err != nil {
		return fmt.Errorf("master key: %w", err)
	}
	st, db, err := server.OpenStore(cfg, kr)
	if err != nil {
		return err
	}
	defer db.Close()
	defer st.Close()
	if st.Empty() {
		return fmt.Errorf("the database %s holds no server yet; start keeporgd first", db.Where)
	}

	password := os.Getenv("KEEPORG_ADMIN_PASSWORD")
	generated := false
	if password == "" {
		password = secret.Token()[:20]
		generated = true
	}
	hash, err := secret.Hash(password)
	if err != nil {
		return err
	}

	var name string
	if err := st.Update(func(s *store.State) error {
		u := s.UserByName("admin")
		if u == nil {
			// Whichever account can administer the server will do.
			for _, candidate := range s.Users {
				if candidate.ServerAdmin {
					u = candidate
					break
				}
			}
		}
		if u == nil {
			u = &store.User{ID: secret.ID(), Username: "admin", CreatedAt: time.Now()}
			s.Users[u.ID] = u
		}
		u.PassHash = hash
		u.TOTPSecret, u.TOTPConfirmed = "", false
		u.ServerAdmin, u.Disabled = true, false
		for token, sess := range s.Sessions {
			if sess.UserID == u.ID {
				delete(s.Sessions, token)
			}
		}
		name = u.Username
		return nil
	}); err != nil {
		return err
	}

	if generated {
		lg.Printf("reset %q: password %q (shown once). Any second factor on that "+
			"account was cleared and its sessions were revoked.", name, password)
	} else {
		lg.Printf("reset %q to the password from KEEPORG_ADMIN_PASSWORD. Any second "+
			"factor on that account was cleared and its sessions were revoked.", name)
	}
	return nil
}

// healthcheck asks a server listening on this configuration whether it is
// serving, and says why not when it is not.
func healthcheck(cfg server.Config) error {
	cfg.Defaults()
	host, port, err := net.SplitHostPort(cfg.Addr)
	if err != nil {
		return fmt.Errorf("address %q: %w", cfg.Addr, err)
	}
	if host == "" || host == "0.0.0.0" || host == "::" {
		host = "127.0.0.1"
	}
	scheme := "http"
	client := &http.Client{Timeout: 5 * time.Second}
	if cfg.TLSCert != "" && cfg.TLSKey != "" {
		scheme = "https"
		// A loopback probe of our own listener: the certificate's name is
		// the public one, which this connection is not using.
		client.Transport = &http.Transport{
			TLSClientConfig: &tls.Config{InsecureSkipVerify: true}, //nolint:gosec
		}
	}
	resp, err := client.Get(scheme + "://" + net.JoinHostPort(host, port) + "/healthz")
	if err != nil {
		return err
	}
	defer resp.Body.Close()
	if resp.StatusCode != http.StatusOK {
		return fmt.Errorf("/healthz returned %s", resp.Status)
	}
	return nil
}

func writeExampleConfig(path string) error {
	ex := fileConfig{
		Addr:            ":8765",
		DataDir:         "/var/lib/keeporg",
		ExternalURL:     "https://keeporg.example.com",
		SessionTTLHours: 24 * 90,
		MaxDBSizeMB:     32,
		Database:        "sqlite:/var/lib/keeporg/keeporg.db",
		TrustedProxies:  []string{"127.0.0.1", "::1"},
		MasterKeyFile:   "/etc/keeporg/master.key",
		VaultKDF:        vaultKDF{MemoryMB: 32, Iterations: 2, Lanes: 2},
		OIDC: server.OIDCConfig{
			Issuer: "https://id.example.com/realms/main", ClientID: "keeporg",
			ClientSecret: "change-me", Scopes: "openid profile email", AutoCreate: false,
		},
	}
	b, err := json.MarshalIndent(ex, "", "  ")
	if err != nil {
		return err
	}
	if err := os.MkdirAll(filepath.Dir(path), 0o755); err != nil {
		return err
	}
	return os.WriteFile(path, append(b, '\n'), 0o600)
}