Files
gohoarder/pkg/storage/nfs/nfs.go
T
lukaszraczylo e39d6a0f0d feat: comprehensive audit + Tier 3 wiring (security/correctness/features)
Multi-agent audit + fix pass covering bugs, security, dead config wiring,
and missing features. All quality gates green: build/vet/test -race/govulncheck/
golangci-lint. Frontend tests 47/47.

SECURITY & CORRECTNESS

- Scanner pipeline fail-closed: cache.go scan timeout now 503 (was goto servePkg);
  scanner.go all-scanners-fail saves ScanStatusError; CheckVulnerabilities blocks
  on missing/error result instead of allowing.
- Scanner argv corrections: govulncheck source-mode + go.mod discovery;
  npm-audit generates lockfile via npm install --package-lock-only --ignore-scripts;
  pip-audit per-extension dispatch (wheel direct / sdist extract+pyproject).
- GHSA version-range filtering implemented (was always-include); url.QueryEscape
  on package name.
- pypi SSRF closed via host allowlist on original_url; url.QueryEscape on
  rewriteURL output.
- Path traversal: cache temp file uses os.CreateTemp; smb keyToPath sanitized;
  filesystem keyToPath returns error + filepath.Abs prefix-verify.
- Goroutine leaks plugged: cache.cleanupWorker stop channel; auth.ValidationCache
  Stop() with sync.Once; ws.unregister buffered with non-blocking send.
- WebSocket double-close panic fixed via sync.Once Client.closeSend().
- Race conditions: gormstore.registryCache sync.RWMutex (was concurrent map
  panic); auth.LastUsedAt no longer mutated under RLock; analytics strict
  lock-ordering invariant (statsMu before downloadsMu).
- Auth validator fails CLOSED on transport/unknown-status errors (was returning
  true,err 'allow cache fallback').
- Credential cache hash full SHA256 (was 8-byte truncate; eliminates 2^32
  collision risk).
- filesystem.fs.used incremented after successful rename (no quota inflation
  race).
- gormstore aggregation events deleted in same tx as insert (no double-counting).
- gormstore.SavePackage uses Updates(map) so zero-value security flags
  (RequiresAuth=false) can be cleared.
- partition_manager validates partition name regex before DROP TABLE.
- vcs/git: version regex validation rejects --option-injection; checkout uses
  --detach -- separator; removed TrimPrefix(repo,'v') that corrupted vault/
  vitess/vim-go.
- WebSocket CheckOrigin allowlist (was return true; CSWSH closed); same-origin
  default + ServerConfig.AllowedOrigins.
- fiber CVE GO-2026-4543 patched (v2.52.10 -> v2.52.12).

FUNCTIONAL FEATURES

- API key DB persistence: APIKeyModel + migration 202604280002, full CRUD,
  async LastUsedAt updates with WaitGroup-tracked goroutines drained on Close.
- Admin bootstrap via GOHOARDER_BOOTSTRAP_ADMIN_KEY env var; idempotent
  (skipped when non-revoked admin already exists).
- Auth middleware factory in pkg/app: RequireAuth, RequireRole, RequirePermission,
  OptionalAuth. Mounted on proxy + read APIs when Auth.Enabled=true.
  DELETE /api/packages/* requires admin role. /health public for k8s probes.
- NFS storage backend: pkg/storage/nfs wraps filesystem with /proc/mounts
  detection (Linux), post-write fsync, read-after-write health probe.
- WebSocket real-time pipeline: pkg/events Broadcaster interface; cache + scanner
  managers emit EventPackageCached / EventPackageDownloaded / EventScanComplete;
  app.go runs 30s EventStatsUpdate ticker. Frontend has full WS client (reconnect
  with exponential backoff 1s->30s, 25s heartbeat, subscriber pattern), Pinia
  realtime store, Dashboard live indicator + activity feed + reactive stats
  override.
- TLS termination: fiber.ListenTLS when Server.TLS.Enabled.
- Pre-warming: Prewarming.Enabled flag wired (was hardcoded false); Interval,
  MaxConcurrent, TopPackages plumbed.
- NetworkConfig wired (timeouts, retry, rate-limit, circuit-breaker).
- AuthConfig.BcryptCost wired.
- Security.Scanners.Static.MaxPackageSize plumbed to cache.io.LimitReader.
- Handlers.{Go,NPM,PyPI}.Enabled gates route mounting.
- Graceful shutdown: authManager.Close drains async writes.

LINT/HYGIENE

- 80 golangci-lint issues resolved: errcheck (Close/Write/RemoveAll wrapped),
  gofmt, gosec G104/G306, govet shadow renames + fieldalignment reorders,
  staticcheck ST1000 pkg comments + QF1003 tagged switches + QF1008 embedded
  field selectors + QF1001 De Morgan's law.

BEHAVIOR CHANGES

- Scanner now fail-closed: deployments without scanner binaries
  (trivy/govulncheck/npm-audit/pip-audit/grype) BLOCK packages instead of
  serving unscanned. Add binaries or plan a future allow-on-scan-error flag.
- WS origin defaults to same-origin; cross-origin dashboards must set
  server.allowed_origins.
- Validator no longer fails open; private packages won't serve from cache when
  validation can't be performed.
- download_events retention dropped from 24h to ~5min (deleted in agg tx).
  Migration 202604280001 purges pre-upgrade events.
- DELETE /api/packages/* requires admin role when auth enabled.

NEW PACKAGES

- pkg/events  - Broadcaster interface
- pkg/storage/nfs  - NFS-aware filesystem wrapper

DEFERRED (out of scope this pass)

- Static scanner implementation (config defines AllowedLicenses/MaxPackageSize/
  BlockSuspicious but pkg/scanner/static/ does not exist).
- AuthConfig.AuditLog wiring (no audit log infra yet).
- CacheConfig.TTLOverrides passthrough (would touch cache.Config beyond
  integrator scope).
2026-07-10 09:37:55 +01:00

281 lines
9.2 KiB
Go

// Package nfs implements an NFS-backed storage backend.
//
// NFS is, from Go's perspective, an ordinary mounted filesystem. The user is
// expected to mount the export at cfg.Path before starting the application;
// this package does NOT perform mount(8) calls. It wraps the filesystem
// backend and adds NFS-specific safety:
//
// - Best-effort mount-type detection (Linux: /proc/mounts). On non-Linux
// platforms detection is skipped silently. A non-NFS mount is logged at
// Warn level but is NOT a fatal error so tests/CI can run on local
// filesystems.
//
// - Optional per-write fsync (SyncWrites, default true) to flush NFS client
// caches and improve durability across NFS-cached metadata. Stale handles
// and "silent" write losses are common NFS pitfalls.
//
// - A richer Health probe that round-trips a marker file (write, fsync,
// read, delete) to surface stale handles or read-after-write
// inconsistencies the bare filesystem health check would miss.
package nfs
import (
"context"
"fmt"
"io"
"os"
"path/filepath"
"runtime"
"strings"
"github.com/lukaszraczylo/gohoarder/pkg/errors"
"github.com/lukaszraczylo/gohoarder/pkg/storage"
"github.com/lukaszraczylo/gohoarder/pkg/storage/filesystem"
"github.com/rs/zerolog"
)
// Config holds NFS storage configuration. The struct is intentionally
// self-contained so callers can map their own config (e.g.
// pkg/config.StorageConfig) without import cycles.
type Config struct {
// Path is the local mount point of the NFS export. Required.
Path string
// MaxSize is the optional quota in bytes (0 = unlimited). Forwarded to
// the underlying filesystem backend.
MaxSize int64
// SyncWrites, when true (default), forces fsync after every successful
// Put so data is flushed through the NFS client cache to the server.
SyncWrites bool
}
// Storage implements storage.StorageBackend on top of an NFS-mounted path.
type Storage struct {
fs *filesystem.FilesystemStorage
logger zerolog.Logger
path string
syncWrites bool
}
// New constructs an NFS storage backend rooted at cfg.Path.
//
// cfg.Path must already exist and be a directory; the caller is responsible
// for the actual NFS mount. Mount-type detection is best-effort.
func New(cfg Config, logger zerolog.Logger) (*Storage, error) {
if cfg.Path == "" {
return nil, errors.New(errors.ErrCodeStorageFailure, "nfs: path is required")
}
info, err := os.Stat(cfg.Path)
if err != nil {
return nil, errors.Wrap(err, errors.ErrCodeStorageFailure, "nfs: path does not exist or is inaccessible")
}
if !info.IsDir() {
return nil, errors.New(errors.ErrCodeStorageFailure, fmt.Sprintf("nfs: path is not a directory: %s", cfg.Path))
}
// Best-effort mount-type detection. Non-fatal: warn only.
if mountType, ok := detectMountType(cfg.Path); ok {
if !isNFSMountType(mountType) {
logger.Warn().
Str("path", cfg.Path).
Str("mount_type", mountType).
Msg("nfs: configured path is not on an NFS mount; proceeding anyway")
} else {
logger.Info().
Str("path", cfg.Path).
Str("mount_type", mountType).
Msg("nfs: detected NFS mount")
}
} else {
// Detection unavailable (non-Linux or /proc/mounts unreadable).
logger.Debug().
Str("path", cfg.Path).
Str("os", runtime.GOOS).
Msg("nfs: mount-type detection skipped")
}
fs, err := filesystem.New(cfg.Path, cfg.MaxSize)
if err != nil {
return nil, err
}
return &Storage{
fs: fs,
logger: logger,
path: cfg.Path,
syncWrites: cfg.SyncWrites,
}, nil
}
// Get delegates to the underlying filesystem backend.
func (s *Storage) Get(ctx context.Context, key string) (io.ReadCloser, error) {
return s.fs.Get(ctx, key)
}
// Put delegates to the filesystem backend and, when SyncWrites is enabled,
// fsyncs the resulting file to flush the NFS client cache.
func (s *Storage) Put(ctx context.Context, key string, data io.Reader, opts *storage.PutOptions) error {
if err := s.fs.Put(ctx, key, data, opts); err != nil {
return err
}
if !s.syncWrites {
return nil
}
// Resolve the on-disk path via the LocalPathProvider contract the
// filesystem backend implements. Failure to fsync is logged but not
// returned: the write itself succeeded; durability is best-effort.
path, err := s.fs.GetLocalPath(ctx, key)
if err != nil {
s.logger.Warn().Err(err).Str("key", key).Msg("nfs: post-put path lookup failed; skipping fsync")
return nil
}
f, err := os.OpenFile(path, os.O_RDWR, 0) // #nosec G304 -- path resolved by sanitizing backend
if err != nil {
s.logger.Warn().Err(err).Str("key", key).Msg("nfs: post-put open failed; skipping fsync")
return nil
}
if syncErr := f.Sync(); syncErr != nil {
s.logger.Warn().Err(syncErr).Str("key", key).Msg("nfs: post-put fsync failed")
}
_ = f.Close() // #nosec G104 -- close after sync, error not actionable
return nil
}
// Delete delegates to the underlying filesystem backend.
func (s *Storage) Delete(ctx context.Context, key string) error {
return s.fs.Delete(ctx, key)
}
// Exists delegates to the underlying filesystem backend.
func (s *Storage) Exists(ctx context.Context, key string) (bool, error) {
return s.fs.Exists(ctx, key)
}
// List delegates to the underlying filesystem backend.
func (s *Storage) List(ctx context.Context, prefix string, opts *storage.ListOptions) ([]storage.StorageObject, error) {
return s.fs.List(ctx, prefix, opts)
}
// Stat delegates to the underlying filesystem backend.
func (s *Storage) Stat(ctx context.Context, key string) (*storage.StorageInfo, error) {
return s.fs.Stat(ctx, key)
}
// GetQuota delegates to the underlying filesystem backend.
func (s *Storage) GetQuota(ctx context.Context) (*storage.QuotaInfo, error) {
return s.fs.GetQuota(ctx)
}
// Health checks both the underlying filesystem and runs an NFS-specific
// round-trip probe (write, fsync, read, delete) to surface stale handles or
// cache-coherency issues that a bare stat would miss.
func (s *Storage) Health(ctx context.Context) error {
if err := s.fs.Health(ctx); err != nil {
return err
}
if err := ctx.Err(); err != nil {
return err
}
probePath := filepath.Join(s.path, ".nfs_health_probe")
payload := []byte("nfs-health-probe")
f, err := os.Create(probePath) // #nosec G304 -- path under configured base, fixed name
if err != nil {
return errors.Wrap(err, errors.ErrCodeStorageFailure, "nfs: cannot create health probe file")
}
if _, writeErr := f.Write(payload); writeErr != nil {
_ = f.Close() // #nosec G104 -- cleanup
_ = os.Remove(probePath) // #nosec G104 -- cleanup
return errors.Wrap(writeErr, errors.ErrCodeStorageFailure, "nfs: cannot write health probe")
}
if syncErr := f.Sync(); syncErr != nil {
_ = f.Close() // #nosec G104 -- cleanup
_ = os.Remove(probePath) // #nosec G104 -- cleanup
return errors.Wrap(syncErr, errors.ErrCodeStorageFailure, "nfs: fsync of health probe failed")
}
if closeErr := f.Close(); closeErr != nil {
_ = os.Remove(probePath) // #nosec G104 -- cleanup
return errors.Wrap(closeErr, errors.ErrCodeStorageFailure, "nfs: close of health probe failed")
}
got, err := os.ReadFile(probePath) // #nosec G304 -- fixed probe path
if err != nil {
_ = os.Remove(probePath) // #nosec G104 -- cleanup
return errors.Wrap(err, errors.ErrCodeStorageFailure, "nfs: read-back of health probe failed (possible stale handle)")
}
if string(got) != string(payload) {
_ = os.Remove(probePath) // #nosec G104 -- cleanup
return errors.New(errors.ErrCodeStorageFailure, "nfs: health probe payload mismatch (cache coherency issue?)")
}
if err := os.Remove(probePath); err != nil {
return errors.Wrap(err, errors.ErrCodeStorageFailure, "nfs: cannot remove health probe file")
}
return nil
}
// Close delegates to the underlying filesystem backend.
func (s *Storage) Close() error {
return s.fs.Close()
}
// GetLocalPath exposes direct on-disk paths for scanning (NFS exports look
// like local files to callers). Implements storage.LocalPathProvider.
func (s *Storage) GetLocalPath(ctx context.Context, key string) (string, error) {
return s.fs.GetLocalPath(ctx, key)
}
// detectMountType returns the filesystem type backing path. Linux-only: on
// other platforms the second return value is false. Implementation walks
// /proc/mounts and selects the longest matching mount point, which is the
// canonical way to find which mount owns a path.
func detectMountType(path string) (string, bool) {
if runtime.GOOS != "linux" {
return "", false
}
abs, err := filepath.Abs(path)
if err != nil {
return "", false
}
data, err := os.ReadFile("/proc/mounts")
if err != nil {
return "", false
}
var (
bestMount string
bestType string
)
for _, line := range strings.Split(string(data), "\n") {
fields := strings.Fields(line)
if len(fields) < 3 {
continue
}
mountPoint := fields[1]
fsType := fields[2]
if abs == mountPoint || strings.HasPrefix(abs, strings.TrimRight(mountPoint, "/")+"/") {
if len(mountPoint) > len(bestMount) {
bestMount = mountPoint
bestType = fsType
}
}
}
if bestMount == "" {
return "", false
}
return bestType, true
}
// isNFSMountType returns true for NFS family mount types.
func isNFSMountType(t string) bool {
switch t {
case "nfs", "nfs4":
return true
default:
return false
}
}