hamlab/cmd/hamlab/main.go

// Ham Lab — Proxmox-style web console for a bare-metal Debian host.
package main

import (
	"bufio"
	"context"
	"encoding/json"
	"errors"
	"flag"
	"fmt"
	"io/fs"
	"log"
	"net"
	"net/http"
	"os"
	"os/signal"
	"os/user"
	"path/filepath"
	"runtime"
	"runtime/debug"
	"strings"
	"sync/atomic"
	"syscall"
	"time"

	"golang.org/x/term"

	"hamlab/internal/api"
	"hamlab/internal/auth"
	"hamlab/internal/config"
	"hamlab/internal/cron"
	"hamlab/internal/files"
	"hamlab/internal/incus"
	"hamlab/internal/media"
	"hamlab/internal/metrics"
	"hamlab/internal/nodes"
	"hamlab/internal/priv"
	"hamlab/internal/services"
	"hamlab/internal/setup"
	"hamlab/internal/space"
	"hamlab/internal/storage"
	"hamlab/internal/store"
	"hamlab/internal/tasks"
	hterm "hamlab/internal/term"
	"hamlab/internal/tlsauto"
	"hamlab/internal/updates"
	"hamlab/internal/web"
)

var version = "dev"

func main() {
	defaultCfg := os.Getenv("HAMLAB_CONFIG")
	if defaultCfg == "" {
		defaultCfg = config.DefaultPath()
	}
	cfgPath := flag.String("config", defaultCfg, "path to config.json")
	flag.Usage = func() {
		fmt.Fprint(os.Stderr, `usage: hamlab [-config path] <command>

  install [flags]      install this binary as a service on this machine (install -h for flags)
  uninstall [--purge]  remove the service and the binary
  serve                run the console in the foreground (what the service runs)
  passwd [user]        set a user's password (makes the account if it is new; reads stdin when piped)
  users                list the accounts
  zfs-snapshot <id>    run one ZFS snapshot schedule (what its cron line calls)
  virt-export [name…]  export Incus instances to the library (what its cron line calls)
  docker-update <id>   run one image-update schedule (what its cron line calls)
  user-rm <user>       remove an account and sign it out everywhere
  check                verify the config and state load, then exit
  version
  push-nodes [--force] hub: send this build to every managed machine

`)
		flag.PrintDefaults()
	}
	flag.Parse()

	cmd := "serve"
	if flag.NArg() > 0 {
		cmd = flag.Arg(0)
	}
	switch cmd { // commands that don't read the config
	case "fs": // privileged helper for the Files tab (run as root by the daemon)
		os.Exit(files.Helper(flag.Args()[1:], os.Stdin, os.Stdout, os.Stderr))
	case "thumb": // privileged helper for the Files tab's previews in root mode (run as root by the daemon)
		os.Exit(media.Helper(flag.Args()[1:], os.Stdin, os.Stdout, os.Stderr))
	case "space-walk": // helper for the Space tab: measure mountpoints, stream the tree back
		os.Exit(space.Helper(flag.Args()[1:], os.Stdin, os.Stdout, os.Stderr))
	case "drm-clients": // privileged helper for the GPU readings: who has the GPU open (run as root by the daemon)
		os.Exit(metrics.DRMClientsHelper(os.Stdin, os.Stdout))
	case "cpu-energy": // privileged helper for the CPU power reading: the packages' energy counters (run as root by the daemon)
		os.Exit(metrics.CPUEnergyHelper(os.Stdin, os.Stdout))
	case "install", "uninstall":
		var err error
		if cmd == "install" {
			err = setup.Install(flag.Args()[1:], version)
		} else {
			err = setup.Uninstall(flag.Args()[1:])
		}
		if errors.Is(err, flag.ErrHelp) {
			os.Exit(0)
		}
		if err != nil {
			fmt.Fprintln(os.Stderr, "hamlab "+cmd+":", err)
			os.Exit(1)
		}
		return
	case "replace": // self-update helper (run as root by the daemon): put this binary at a path, atomically
		// Only ever over a hamlab binary: this runs as root, and must not be a way to
		// overwrite arbitrary files for whoever is allowed to run hamlab as root.
		if flag.NArg() != 2 || !filepath.IsAbs(flag.Arg(1)) || !strings.HasPrefix(filepath.Base(flag.Arg(1)), "hamlab") {
			fmt.Fprintln(os.Stderr, "hamlab replace: destination must be an absolute path to a hamlab binary")
			os.Exit(2)
		}
		self, err := os.Executable()
		if err == nil {
			err = setup.CopyAtomic(self, flag.Arg(1))
		}
		if err != nil {
			fmt.Fprintln(os.Stderr, "hamlab replace:", err)
			os.Exit(1)
		}
		return
	}
	cfg, err := config.Load(*cfgPath)
	if err != nil {
		log.Fatal(err)
	}
	switch cmd {
	case "version":
		fmt.Println("hamlab", version)
	case "check": // a new build runs this against the live config before it replaces the old one
		if _, err := store.Open(filepath.Join(cfg.DataDir(), "state.json")); err != nil {
			log.Fatal(err)
		}
		if _, err := fs.Stat(web.FS(), "index.html"); err != nil {
			log.Fatal("embedded UI missing: ", err)
		}
		fmt.Println("ok")
	case "passwd":
		if err := passwd(cfg, flag.Arg(1)); err != nil {
			log.Fatal(err)
		}
	case "docker-update":
		if err := dockerUpdateJob(cfg, flag.Arg(1)); err != nil {
			log.Fatal(err)
		}
	case "zfs-snapshot":
		if err := zfsSnapshot(cfg, flag.Arg(1)); err != nil {
			log.Fatal(err)
		}
	case "virt-export": // the nightly cron line from the Virtualization tab; optional instance names
		if err := virtExport(cfg, flag.Args()[1:]); err != nil {
			log.Fatal(err)
		}
	case "users":
		migrateUsers(cfg)
		if len(cfg.Users) == 0 {
			fmt.Println("no accounts in", cfg.Path())
		}
		for _, u := range cfg.Users {
			seen := "never"
			if !u.LastSeen.IsZero() {
				seen = u.LastSeen.Local().Format("2006-01-02 15:04")
			}
			note := ""
			if u.Guest { // a temporary look-only login, not one of the machine's accounts
				note = "  guest, until " + u.Expires.Local().Format("15:04")
				if u.Expired(time.Now()) {
					note = "  guest, expired"
				}
			}
			fmt.Printf("%-20s created %s  last seen %s%s\n", u.Name, u.Created.Local().Format("2006-01-02"), seen, note)
		}
	case "user-rm":
		if err := userRm(cfg, flag.Arg(1)); err != nil {
			log.Fatal(err)
		}
	case "push-nodes": // hub CLI: send this binary to every managed machine
		exe, _ := os.Executable()
		host, _ := os.Hostname()
		nm := nodes.New(cfg, host)
		force := flag.NArg() > 1 && flag.Arg(1) == "--force"
		results := nm.PushAll(context.Background(), exe, runtime.GOARCH, force)
		if len(results) == 0 {
			fmt.Println("no managed machines in", cfg.Path())
		}
		failed := 0
		for _, r := range results {
			if r.OK {
				fmt.Printf("%-12s updated to %s\n", r.Name, r.Version)
			} else {
				failed++
				fmt.Printf("%-12s FAILED: %s\n", r.Name, r.Error)
			}
		}
		if failed > 0 {
			os.Exit(1)
		}
	case "serve":
		if err := serve(cfg); err != nil {
			log.Fatal(err)
		}
	default:
		flag.Usage()
		os.Exit(2)
	}
}

// migrateUsers gives a pre-users config its first account, named after this OS user, so
// the CLI sees the same accounts the daemon will.
// dockerUpdateJob pulls a stack's images, and deploys them when the schedule says to. This
// is what the cron line Ham Lab writes actually calls, so image updates keep happening
// whether or not the console is up.
func dockerUpdateJob(cfg *config.Config, id string) error {
	if strings.TrimSpace(id) == "" {
		return errors.New("usage: hamlab docker-update <schedule-id>")
	}
	st, err := store.Open(filepath.Join(cfg.DataDir(), "state.json"))
	if err != nil {
		return err
	}
	var raw json.RawMessage
	st.View(func(s *store.State) { raw = s.DockerJobs })
	var d struct {
		Jobs []api.UpdateJob `json:"jobs"`
	}
	if len(raw) > 0 {
		if err := json.Unmarshal(raw, &d); err != nil {
			return fmt.Errorf("reading the schedules: %w", err)
		}
	}
	for _, j := range d.Jobs {
		if j.ID != id {
			continue
		}
		// A pull that never finishes must not leave a cron job resident until the next reboot.
		ctx, cancel := context.WithTimeout(context.Background(), 2*time.Hour)
		defer cancel()
		out, runErr := api.RunUpdate(ctx, st, j, api.DockerTool())
		rec := storage.RunRecord{At: time.Now(), Taken: j.Stack}
		if runErr != nil {
			rec.Error = runErr.Error()
		}
		if err := storage.WriteRun(cfg.DataDir(), "docker-"+id, rec); err != nil {
			fmt.Fprintln(os.Stderr, "recording the run:", err)
		}
		fmt.Print(out)
		return runErr
	}
	return fmt.Errorf("no image-update schedule %q — it may have been removed; this cron line is stale", id)
}

// zfsSnapshot runs one snapshot schedule: take it, then expire what has aged out. This is
// what the cron line Ham Lab writes actually calls, so a schedule keeps running whether or
// not the console is up — and keeps running if Ham Lab is never opened again.
//
// It deliberately does not write state.json. The daemon owns that file and would clobber
// anything written from out here; the outcome goes beside it instead, in zfs-runs/.
func zfsSnapshot(cfg *config.Config, id string) error {
	if strings.TrimSpace(id) == "" {
		return errors.New("usage: hamlab zfs-snapshot <schedule-id>")
	}
	st, err := store.Open(filepath.Join(cfg.DataDir(), "state.json"))
	if err != nil {
		return err
	}
	var raw json.RawMessage
	st.View(func(s *store.State) { raw = s.ZFS })
	var z struct {
		Policies []storage.Policy `json:"policies"`
	}
	if len(raw) > 0 {
		if err := json.Unmarshal(raw, &z); err != nil {
			return fmt.Errorf("reading the schedules: %w", err)
		}
	}
	var p *storage.Policy
	for i := range z.Policies {
		if z.Policies[i].ID == id {
			p = &z.Policies[i]
		}
	}
	if p == nil {
		return fmt.Errorf("no snapshot schedule %q — it may have been removed; this cron line is stale", id)
	}
	res, runErr := storage.New(priv.Available()).Run(context.Background(), p)
	rec := storage.RunRecord{At: time.Now(), Taken: res.Taken, Pruned: res.Pruned}
	if runErr != nil {
		rec.Error = runErr.Error()
	}
	if err := storage.WriteRun(cfg.DataDir(), id, rec); err != nil {
		fmt.Fprintln(os.Stderr, "recording the run:", err)
	}
	if runErr != nil {
		return runErr
	}
	fmt.Printf("%s: took %s", p.Name, res.Taken)
	if len(res.Pruned) > 0 {
		fmt.Printf(", expired %d", len(res.Pruned))
	}
	fmt.Println()
	return nil
}

// virtExport runs the instance exports the Virtualization tab scheduled, as the account Ham
// Lab runs as (which is in incus-admin), and records the run beside the data.
func virtExport(cfg *config.Config, names []string) error {
	who := os.Getenv("USER")
	if u, err := user.Current(); err == nil && u.Username != "" {
		who = u.Username
	}
	run := incus.New(cfg.DataDir(), who, version).RunExports(context.Background(), names, func(l string) { fmt.Println(l) })
	fmt.Printf("exported %d, pruned %d, %.0f s\n", len(run.Exported), len(run.Pruned), run.Seconds)
	if len(run.Errors) > 0 {
		return errors.New(strings.Join(run.Errors, "; "))
	}
	return nil
}

func migrateUsers(cfg *config.Config) {
	name := os.Getenv("USER")
	if u, err := user.Current(); err == nil && u.Username != "" {
		name = u.Username
	}
	if cfg.MigrateUsers(name) {
		_ = cfg.Save()
	}
}

func userRm(cfg *config.Config, name string) error {
	migrateUsers(cfg)
	if name == "" {
		return errors.New("usage: hamlab user-rm <user>")
	}
	if len(cfg.Users) <= 1 {
		return errors.New("the last account can't be removed")
	}
	kept := cfg.Users[:0]
	found := false
	for _, u := range cfg.Users {
		if u.Name == name {
			found = true
			continue
		}
		kept = append(kept, u)
	}
	if !found {
		return fmt.Errorf("no account %q", name)
	}
	cfg.Users = kept
	if err := cfg.Save(); err != nil {
		return err
	}
	if st, err := store.Open(filepath.Join(cfg.DataDir(), "state.json")); err == nil {
		auth.NewSessions(st).RevokeUser(name)
		_ = st.Update(func(s *store.State) { delete(s.Users, name) })
	}
	fmt.Fprintln(os.Stderr, "removed", name, "(the running service picks it up at its next config save; restart it to be sure)")
	return nil
}

func passwd(cfg *config.Config, name string) error {
	migrateUsers(cfg)
	if name == "" {
		switch len(cfg.Users) {
		case 0:
			name = os.Getenv("USER")
			if u, err := user.Current(); err == nil && u.Username != "" {
				name = u.Username
			}
		case 1:
			name = cfg.Users[0].Name
		default:
			return errors.New("several accounts: hamlab passwd <user>")
		}
	}
	if !auth.UserNameRe.MatchString(name) {
		return fmt.Errorf("%q: a user name is 1–32 lower-case letters, digits, . _ or -", name)
	}
	var p1 []byte
	if !term.IsTerminal(int(os.Stdin.Fd())) { // scripted: one line on stdin
		line, _ := bufio.NewReader(os.Stdin).ReadString('\n')
		p1 = []byte(strings.TrimRight(line, "\r\n"))
	} else {
		fmt.Fprintf(os.Stderr, "New password for %s: ", name)
		var err error
		p1, err = term.ReadPassword(int(os.Stdin.Fd()))
		fmt.Fprintln(os.Stderr)
		if err != nil {
			return err
		}
		fmt.Fprint(os.Stderr, "Again: ")
		p2, err := term.ReadPassword(int(os.Stdin.Fd()))
		fmt.Fprintln(os.Stderr)
		if err != nil {
			return err
		}
		if string(p1) != string(p2) {
			return errors.New("passwords do not match")
		}
	}
	if len(p1) < auth.MinPasswordLen {
		return errors.New("password must be at least 12 characters")
	}
	h, err := auth.HashPassword(string(p1))
	if err != nil {
		return err
	}
	made := true
	for i := range cfg.Users {
		if cfg.Users[i].Name == name {
			cfg.Users[i].Hash = h
			made = false
		}
	}
	if made {
		cfg.Users = append(cfg.Users, config.User{Name: name, Hash: h, Created: time.Now()})
	}
	if err := cfg.Save(); err != nil {
		return err
	}
	if made {
		fmt.Fprintln(os.Stderr, "account", name, "made in", cfg.Path())
	} else {
		fmt.Fprintln(os.Stderr, "password of", name, "set in", cfg.Path())
	}
	fmt.Fprintln(os.Stderr, "the running service reads the config at its next save; restart it to be sure")
	return nil
}

func serve(cfg *config.Config) error {
	if cfg.PasswordHash == "" && len(cfg.Users) == 0 {
		return fmt.Errorf("no account set — run: hamlab -config %s passwd", cfg.Path())
	}
	if err := os.MkdirAll(cfg.DataDir(), 0o700); err != nil {
		return err
	}
	st, err := store.Open(filepath.Join(cfg.DataDir(), "state.json"))
	if err != nil {
		return err
	}
	sessions := auth.NewSessions(st)

	u, err := user.Current()
	if err != nil {
		return err
	}
	shell := os.Getenv("SHELL")
	if shell == "" {
		shell = loginShell(u.Username)
	}
	mgr := hterm.NewManager(shell, u.HomeDir)
	mgr.OnChange = func(list []hterm.Info) {
		_ = st.Update(func(s *store.State) {
			s.Terminals = s.Terminals[:0]
			for _, t := range list {
				if t.Running { // an exited shell must not come back from the dead on restart
					s.Terminals = append(s.Terminals, store.Terminal{ID: t.ID, Name: t.Name})
				}
			}
		})
	}
	// Re-open the shells that existed before the last restart (fresh processes,
	// same names/order) so the sidebar comes back as it was.
	var saved []store.Terminal
	st.View(func(s *store.State) { saved = append(saved, s.Terminals...) })
	for _, t := range saved {
		if _, err := mgr.Create(t.ID, t.Name); err != nil {
			log.Printf("restore terminal %q: %v", t.Name, err)
		}
	}

	// Keep the heap tidy: the daemon's live data is tens of MB; without a limit the GC
	// happily lets RSS coast at twice that. (Soft limit: exceeding it only means more GC.)
	debug.SetMemoryLimit(160 << 20)

	col := metrics.NewCollector()
	go col.Run(context.Background())

	svc := services.New()
	cr := cron.New(svc.Sudo(), svc.JournalSudo())
	tm := tasks.New()
	up := updates.New(svc.Sudo(), tm)
	sto := storage.New(svc.Sudo())
	self, _ := os.Executable()
	var sudoFS files.FS
	if svc.Sudo() && self != "" {
		sudoFS = files.Sudo{Self: self}
	}
	host, _ := os.Hostname()
	nm := nodes.New(cfg, host)
	go nm.Run(context.Background())

	certPath, keyPath, fingerprint := "", "", ""
	if cfg.TLS == "auto" {
		var err error
		certPath, keyPath, fingerprint, err = tlsauto.Ensure(cfg.DataDir())
		if err != nil {
			return fmt.Errorf("tls: %w", err)
		}
	}
	ln, err := listen(cfg.Listen)
	if err != nil {
		return err
	}
	srv := api.New(cfg, sessions, mgr, col, st, svc, cr, tm, up, sto, files.Local{}, sudoFS, nm, fingerprint, web.FS(), version)
	hs := &http.Server{
		Handler:           srv.Handler(),
		ReadHeaderTimeout: 10 * time.Second,
	}
	// After a self-update: become the new binary in place. Same PID, so it works the same
	// under systemd, OpenRC, or no supervisor at all. It returns only if the exec failed, by
	// which point the listener and the shells are gone; the caller puts the old binary back
	// and exits so the supervisor starts it.
	var restarting atomic.Bool
	srv.Restart = func() error {
		restarting.Store(true) // Serve returns once Shutdown is called; serve() must not exit under us
		ctx, cancel := context.WithTimeout(context.Background(), 3*time.Second)
		defer cancel()
		_ = hs.Shutdown(ctx)
		mgr.Shutdown()
		return syscall.Exec(self, os.Args, os.Environ())
	}
	go func() {
		sig := make(chan os.Signal, 1)
		signal.Notify(sig, syscall.SIGINT, syscall.SIGTERM)
		<-sig
		log.Print("shutting down")
		ctx, cancel := context.WithTimeout(context.Background(), 5*time.Second)
		defer cancel()
		_ = hs.Shutdown(ctx)
		mgr.Shutdown()
	}()
	if certPath != "" {
		log.Printf("hamlab %s listening on https://%s as %s (shell %s) cert %s", version, cfg.Listen, u.Username, shell, fingerprint)
		if err := hs.ServeTLS(ln, certPath, keyPath); !errors.Is(err, http.ErrServerClosed) {
			return err
		}
		return waitIf(&restarting)
	}
	log.Printf("hamlab %s listening on %s as %s (shell %s)", version, cfg.Listen, u.Username, shell)
	if err := hs.Serve(ln); !errors.Is(err, http.ErrServerClosed) {
		return err
	}
	return waitIf(&restarting)
}

// waitIf parks the caller while a restart is replacing the process.
func waitIf(restarting *atomic.Bool) error {
	if restarting.Load() {
		select {}
	}
	return nil
}

// listen binds the configured address with IP_FREEBIND, so a LAN address that isn't up yet
// (boot, DHCP still negotiating, cable out) doesn't stop the daemon: it binds now and
// answers as soon as the address appears.
func listen(addr string) (net.Listener, error) {
	lc := net.ListenConfig{Control: func(network, _ string, c syscall.RawConn) error {
		return c.Control(func(fd uintptr) {
			_ = syscall.SetsockoptInt(int(fd), syscall.IPPROTO_IP, syscall.IP_FREEBIND, 1)
		})
	}}
	ln, err := lc.Listen(context.Background(), "tcp", addr)
	if err != nil {
		return nil, fmt.Errorf("listen %s: %w", addr, err)
	}
	if host, _, _ := net.SplitHostPort(addr); host != "" {
		if ip := net.ParseIP(host); ip != nil && !ip.IsUnspecified() && !ip.IsLoopback() && !onInterface(ip) {
			log.Printf("listen address %s isn't on any interface yet; bound anyway and waiting for it", host)
		}
	}
	return ln, nil
}

func onInterface(ip net.IP) bool {
	addrs, _ := net.InterfaceAddrs()
	for _, a := range addrs {
		if n, ok := a.(*net.IPNet); ok && n.IP.Equal(ip) {
			return true
		}
	}
	return false
}

// loginShell returns the user's shell from /etc/passwd, falling back to bash if
// present and /bin/sh otherwise. Systemd units don't export $SHELL.
func loginShell(username string) string {
	if b, err := os.ReadFile("/etc/passwd"); err == nil {
		for _, line := range strings.Split(string(b), "\n") {
			f := strings.Split(line, ":")
			if len(f) >= 7 && f[0] == username && f[6] != "" {
				if _, err := os.Stat(f[6]); err == nil {
					return f[6]
				}
			}
		}
	}
	for _, sh := range []string{"/bin/bash", "/usr/bin/bash", "/bin/zsh", "/bin/sh"} {
		if _, err := os.Stat(sh); err == nil {
			return sh
		}
	}
	return "/bin/sh"
}