Compare commits

..
Author SHA1 Message Date
sneak 927fc05cfa Refuse credentials in a storage URL's endpoint or rclone remote name (closes #322)
check / check (push) Waiting to run
storage_url refused only the outer URL's userinfo and unknown query
parameters. Userinfo inside the s3 endpoint= value, and rclone's
remote,param=value parameters in the remote name, were accepted and
then printed by info, remote info and snapshot create. Both are now
refused without echoing the value: a host name never contains "@", and
an rclone remote name never contains a comma. The comma is checked
before url.Parse, because url.Parse refuses a name such as
":s3,secret_access_key=..." or "x,secret=a:b", and its error prints
the name from the colon on. The rclone case has its own error pointing
at rclone config, since the existing credentials error names the s3.*
keys.

Model: opus-5-5
2026-10-13 14:43:40 +00:00
15 changed files with 242 additions and 151 deletions
+6 -1
View File
@@ -58,7 +58,12 @@ func (r *BlobChunkRepository) GetByBlobID(
return nil, fmt.Errorf("querying blob chunks: %w", err)
}
defer func() { _ = rows.Close() }()
defer func() {
err := rows.Close()
if err != nil {
Fatalf("failed to close rows: %v", err)
}
}()
var blobChunks []*BlobChunk
+6 -1
View File
@@ -83,7 +83,12 @@ func (r *BlobRepository) GetAll(ctx context.Context) (map[string]*Blob, error) {
return nil, fmt.Errorf("querying blobs: %w", err)
}
defer func() { _ = rows.Close() }()
defer func() {
err := rows.Close()
if err != nil {
Fatalf("failed to close rows: %v", err)
}
}()
out := make(map[string]*Blob)
+18 -3
View File
@@ -61,7 +61,12 @@ func (r *ChunkFileRepository) GetByChunkHash(
return nil, fmt.Errorf("querying chunk files: %w", err)
}
defer func() { _ = rows.Close() }()
defer func() {
err := rows.Close()
if err != nil {
Fatalf("failed to close rows: %v", err)
}
}()
return r.scanChunkFiles(rows)
}
@@ -82,7 +87,12 @@ func (r *ChunkFileRepository) GetByFilePath(
return nil, fmt.Errorf("querying chunk files: %w", err)
}
defer func() { _ = rows.Close() }()
defer func() {
err := rows.Close()
if err != nil {
Fatalf("failed to close rows: %v", err)
}
}()
return r.scanChunkFiles(rows)
}
@@ -102,7 +112,12 @@ func (r *ChunkFileRepository) GetByFileID(
return nil, fmt.Errorf("querying chunk files: %w", err)
}
defer func() { _ = rows.Close() }()
defer func() {
err := rows.Close()
if err != nil {
Fatalf("failed to close rows: %v", err)
}
}()
return r.scanChunkFiles(rows)
}
+12 -2
View File
@@ -107,7 +107,12 @@ func (r *ChunkRepository) GetByHashes(
return nil, fmt.Errorf("querying chunks: %w", err)
}
defer func() { _ = rows.Close() }()
defer func() {
err := rows.Close()
if err != nil {
Fatalf("failed to close rows: %v", err)
}
}()
var chunks []*Chunk
@@ -147,7 +152,12 @@ func (r *ChunkRepository) ListUnpacked(
return nil, fmt.Errorf("querying unpacked chunks: %w", err)
}
defer func() { _ = rows.Close() }()
defer func() {
err := rows.Close()
if err != nil {
Fatalf("failed to close rows: %v", err)
}
}()
var chunks []*Chunk
+6 -1
View File
@@ -38,7 +38,12 @@ func (r *ChunkRepository) list(ctx context.Context, query string) ([]*Chunk, err
return nil, fmt.Errorf("querying chunks: %w", err)
}
defer func() { _ = rows.Close() }()
defer func() {
err := rows.Close()
if err != nil {
Fatalf("failed to close rows: %v", err)
}
}()
var chunks []*Chunk
+12
View File
@@ -0,0 +1,12 @@
package database
import (
"fmt"
"os"
)
// Fatalf prints an error message to stderr and exits with status 1
func Fatalf(format string, args ...any) {
fmt.Fprintf(os.Stderr, "FATAL: "+format+"\n", args...)
os.Exit(1)
}
+18 -3
View File
@@ -62,7 +62,12 @@ func (r *FileChunkRepository) GetByPath(
return nil, fmt.Errorf("querying file chunks: %w", err)
}
defer func() { _ = rows.Close() }()
defer func() {
err := rows.Close()
if err != nil {
Fatalf("failed to close rows: %v", err)
}
}()
return r.scanFileChunks(rows)
}
@@ -83,7 +88,12 @@ func (r *FileChunkRepository) GetByFileID(
return nil, fmt.Errorf("querying file chunks: %w", err)
}
defer func() { _ = rows.Close() }()
defer func() {
err := rows.Close()
if err != nil {
Fatalf("failed to close rows: %v", err)
}
}()
return r.scanFileChunks(rows)
}
@@ -107,7 +117,12 @@ func (r *FileChunkRepository) GetByPathTx(
return nil, fmt.Errorf("querying file chunks: %w", err)
}
defer func() { _ = rows.Close() }()
defer func() {
err := rows.Close()
if err != nil {
Fatalf("failed to close rows: %v", err)
}
}()
fileChunks, err := r.scanFileChunks(rows)
LogSQL("GetByPathTx", "Complete", path, "count", len(fileChunks))
+24 -4
View File
@@ -175,7 +175,12 @@ func (r *FileRepository) ListModifiedSince(
return nil, fmt.Errorf("querying files: %w", err)
}
defer func() { _ = rows.Close() }()
defer func() {
err := rows.Close()
if err != nil {
Fatalf("failed to close rows: %v", err)
}
}()
var files []*File
@@ -251,7 +256,12 @@ func (r *FileRepository) ListUnderPath(
return nil, fmt.Errorf("querying files: %w", err)
}
defer func() { _ = rows.Close() }()
defer func() {
err := rows.Close()
if err != nil {
Fatalf("failed to close rows: %v", err)
}
}()
var files []*File
@@ -294,7 +304,12 @@ func (r *FileRepository) ListIDsWithChunksNotInUploadedBlobs(
return nil, fmt.Errorf("querying files: %w", err)
}
defer func() { _ = rows.Close() }()
defer func() {
err := rows.Close()
if err != nil {
Fatalf("failed to close rows: %v", err)
}
}()
var ids []types.FileID
@@ -325,7 +340,12 @@ func (r *FileRepository) ListAll(ctx context.Context) ([]*File, error) {
return nil, fmt.Errorf("querying files: %w", err)
}
defer func() { _ = rows.Close() }()
defer func() {
err := rows.Close()
if err != nil {
Fatalf("failed to close rows: %v", err)
}
}()
var files []*File
+9 -11
View File
@@ -106,24 +106,22 @@ func (r *Repositories) WithReadTx(ctx context.Context, fn TxFunc) error {
defer func() {
if p := recover(); p != nil {
// The panic is the failure to report, so a rollback error
// is dropped.
_ = tx.Rollback()
rollbackErr := tx.Rollback()
if rollbackErr != nil {
Fatalf("failed to rollback transaction: %v", rollbackErr)
}
panic(p)
} else if err != nil {
rollbackErr := tx.Rollback()
if rollbackErr != nil {
Fatalf("failed to rollback transaction: %v", rollbackErr)
}
}
}()
err = fn(ctx, tx)
if err != nil {
// Once ctx is cancelled, database/sql rolls the transaction
// back itself and Rollback returns sql.ErrTxDone.
rollbackErr := tx.Rollback()
if rollbackErr != nil && !errors.Is(rollbackErr, sql.ErrTxDone) {
return fmt.Errorf("%w; rolling back transaction: %w",
err, rollbackErr)
}
return err
}
-38
View File
@@ -411,41 +411,3 @@ func TestRepositoriesReadTransaction(t *testing.T) {
t.Error("expected to retrieve file in read transaction")
}
}
// When the context is cancelled during a read transaction's body,
// database/sql rolls the transaction back itself, so WithReadTx's own
// rollback finds it already done. WithReadTx must return the body's error
// alone, without that rollback.
func TestRepositoriesReadTransactionContextCancelled(t *testing.T) {
t.Parallel()
db, cleanup := setupTestDB(t)
defer cleanup()
ctx, cancel := context.WithCancel(context.Background())
defer cancel()
repos := database.NewRepositories(db)
err := repos.WithReadTx(ctx, func(ctx context.Context, tx *sql.Tx) error {
cancel()
// database/sql rolls back from a goroutine of its own; wait
// until it has, so the test does not depend on that race.
for {
_, err := tx.ExecContext(context.WithoutCancel(ctx), "SELECT 1")
if errors.Is(err, sql.ErrTxDone) {
return ctx.Err()
}
time.Sleep(time.Millisecond)
}
})
if !errors.Is(err, context.Canceled) {
t.Fatalf("expected context.Canceled, got: %v", err)
}
if errors.Is(err, sql.ErrTxDone) {
t.Fatalf("the already-done rollback was reported: %v", err)
}
}
+30 -5
View File
@@ -239,7 +239,12 @@ func (r *SnapshotRepository) GetOnlySnapshot(ctx context.Context) (*Snapshot, er
return nil, fmt.Errorf("querying snapshots: %w", err)
}
defer func() { _ = rows.Close() }()
defer func() {
err := rows.Close()
if err != nil {
Fatalf("failed to close rows: %v", err)
}
}()
snapshots, err := r.scanSnapshotRows(rows)
if err != nil {
@@ -274,7 +279,12 @@ func (r *SnapshotRepository) ListRecent(
return nil, fmt.Errorf("querying snapshots: %w", err)
}
defer func() { _ = rows.Close() }()
defer func() {
err := rows.Close()
if err != nil {
Fatalf("failed to close rows: %v", err)
}
}()
return r.scanSnapshotRows(rows)
}
@@ -489,7 +499,12 @@ func (r *SnapshotRepository) GetBlobHashes(
return nil, fmt.Errorf("querying blob hashes: %w", err)
}
defer func() { _ = rows.Close() }()
defer func() {
err := rows.Close()
if err != nil {
Fatalf("failed to close rows: %v", err)
}
}()
var blobs []string
@@ -638,7 +653,12 @@ func (r *SnapshotRepository) GetIncompleteSnapshots(
return nil, fmt.Errorf("querying incomplete snapshots: %w", err)
}
defer func() { _ = rows.Close() }()
defer func() {
err := rows.Close()
if err != nil {
Fatalf("failed to close rows: %v", err)
}
}()
return r.scanSnapshotRows(rows)
}
@@ -661,7 +681,12 @@ func (r *SnapshotRepository) GetIncompleteByHostname(
return nil, fmt.Errorf("querying incomplete snapshots: %w", err)
}
defer func() { _ = rows.Close() }()
defer func() {
err := rows.Close()
if err != nil {
Fatalf("failed to close rows: %v", err)
}
}()
// Same column set as every other multi-row snapshot query, so the
// shared scanner applies — including its timestamp normalization.
-62
View File
@@ -1,62 +0,0 @@
//nolint:testpackage // exercises the unexported cleanupBlobTempFile
package snapshot
import (
"errors"
"os"
"testing"
"github.com/spf13/afero"
"sneak.berlin/go/vaultik/internal/blob"
)
// A blob temp file that cannot be closed or removed must make the cleanup
// return an error, not end the process.
func TestCleanupBlobTempFileReturnsErrors(t *testing.T) {
t.Parallel()
fs := afero.NewOsFs()
s := &Scanner{fs: fs}
tests := []struct {
name string
// prepare runs before the cleanup and makes its close or remove
// fail.
prepare func(tempFile afero.File) error
want error
}{
{
name: "close fails",
prepare: func(tempFile afero.File) error { return tempFile.Close() },
want: os.ErrClosed,
},
{
name: "remove fails",
prepare: func(tempFile afero.File) error {
return fs.Remove(tempFile.Name())
},
want: os.ErrNotExist,
},
}
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
t.Parallel()
tempFile, err := afero.TempFile(fs, t.TempDir(), "blob-*.tmp")
if err != nil {
t.Fatal(err)
}
err = tt.prepare(tempFile)
if err != nil {
t.Fatal(err)
}
err = s.cleanupBlobTempFile(&blob.WithReader{TempFile: tempFile})
if !errors.Is(err, tt.want) {
t.Fatalf("expected an error wrapping %v, got: %v", tt.want, err)
}
})
}
}
+8 -14
View File
@@ -1553,24 +1553,20 @@ func (s *Scanner) handleBlobReady(
blobExists, err := s.uploadBlobIfNeeded(
ctx, blobPath, blobWithReader, startTime, result)
if err != nil {
cleanupErr := s.cleanupBlobTempFile(blobWithReader)
s.cleanupBlobTempFile(blobWithReader)
return errors.Join(
fmt.Errorf("uploading blob %s: %w", finishedBlob.Hash, err), cleanupErr)
return fmt.Errorf("uploading blob %s: %w", finishedBlob.Hash, err)
}
err = s.recordBlobMetadata(ctx, finishedBlob, blobExists, startTime)
if err != nil {
cleanupErr := s.cleanupBlobTempFile(blobWithReader)
s.cleanupBlobTempFile(blobWithReader)
return errors.Join(err, cleanupErr)
}
err = s.cleanupBlobTempFile(blobWithReader)
if err != nil {
return err
}
s.cleanupBlobTempFile(blobWithReader)
// Chunks from this blob are now committed to DB - remove from pending set
s.removePendingChunkHashes(blobWithReader.InsertedChunkHashes)
@@ -1763,22 +1759,20 @@ func (s *Scanner) recordBlobMetadata(
}
// cleanupBlobTempFile closes and removes the blob's temporary file
func (s *Scanner) cleanupBlobTempFile(blobWithReader *blob.WithReader) error {
func (s *Scanner) cleanupBlobTempFile(blobWithReader *blob.WithReader) {
if blobWithReader.TempFile != nil {
tempName := blobWithReader.TempFile.Name()
err := blobWithReader.TempFile.Close()
if err != nil {
return fmt.Errorf("closing blob temp file: %w", err)
log.Fatal("Failed to close temp file", "file", tempName, "error", err)
}
err = s.fs.Remove(tempName)
if err != nil {
return fmt.Errorf("removing blob temp file: %w", err)
log.Fatal("Failed to remove temp file", "file", tempName, "error", err)
}
}
return nil
}
// streamingChunkInfo tracks chunk metadata collected during streaming
+32 -6
View File
@@ -29,7 +29,10 @@ var (
ErrURLCredentials = errors.New(
"storage URL must not carry credentials; " +
"set s3.access_key_id and s3.secret_access_key in the config instead")
ErrURLUnknownParam = errors.New("unknown query parameter in storage URL")
ErrURLUnknownParam = errors.New("unknown query parameter in storage URL")
ErrRcloneRemoteParams = errors.New(
"rclone URL must not carry parameters in the remote name; " +
"set them on the remote with rclone config instead")
)
// URL represents a parsed storage URL.
@@ -90,8 +93,9 @@ func parseFileURL(rawPath string) (*URL, error) {
}
// parseS3URL parses an s3://bucket/prefix URL. It rejects credentials in
// the userinfo and any query parameter other than endpoint, region and
// ssl, so a credential-bearing URL is never stored or echoed.
// the userinfo or in the endpoint parameter's userinfo, and any query
// parameter other than endpoint, region and ssl, so a credential-bearing
// URL is never stored or echoed.
func parseS3URL(rawURL string) (*URL, error) {
u, err := url.Parse(rawURL)
if err != nil {
@@ -114,6 +118,14 @@ func parseS3URL(rawURL string) (*URL, error) {
return nil, err
}
// A host name never contains "@", so one in an endpoint, with or
// without a scheme, can only introduce userinfo.
for _, endpoint := range query["endpoint"] {
if strings.Contains(endpoint, "@") {
return nil, ErrURLCredentials
}
}
return &URL{
Scheme: schemeS3,
Bucket: bucket,
@@ -125,9 +137,23 @@ func parseS3URL(rawURL string) (*URL, error) {
}
// parseRcloneURL parses an rclone://remote/path URL. rclone:// takes no
// query parameters, so credentials in the userinfo and any parameter at
// all are rejected rather than silently ignored.
// query parameters, so credentials in the userinfo, parameters in the
// remote name and any query parameter at all are rejected rather than
// silently ignored.
func parseRcloneURL(rawURL string) (*URL, error) {
// The remote name runs up to the first "/", "?" or "#". It never
// contains a comma; one starts rclone's remote,param=value form,
// whose values can be credentials. This is checked before url.Parse,
// whose error for such a name can print part of it.
remote := strings.TrimPrefix(rawURL, "rclone://")
if end := strings.IndexAny(remote, "/?#"); end >= 0 {
remote = remote[:end]
}
if strings.Contains(remote, ",") {
return nil, ErrRcloneRemoteParams
}
u, err := url.Parse(rawURL)
if err != nil {
return nil, wrapParseError(err)
@@ -137,7 +163,7 @@ func parseRcloneURL(rawURL string) (*URL, error) {
return nil, ErrURLCredentials
}
remote := u.Host
remote = u.Host
if remote == "" {
return nil, ErrMissingRemote
}
+61
View File
@@ -150,6 +150,19 @@ func TestParseStorageURLRejectsCredentials(t *testing.T) {
wantErr: storage.ErrURLCredentials,
secrets: []string{key, secret, "mybucket"},
},
{
name: "s3 userinfo in endpoint",
raw: "s3://mybucket?endpoint=https://" +
key + ":" + secret + "@minio.example.com",
wantErr: storage.ErrURLCredentials,
secrets: []string{key, secret},
},
{
name: "s3 userinfo in endpoint without scheme",
raw: "s3://mybucket?endpoint=" + key + ":" + secret + "@minio.example.com",
wantErr: storage.ErrURLCredentials,
secrets: []string{key, secret},
},
{
name: "s3 unknown query param",
raw: "s3://mybucket?access_key=" + key + "&secret=" + secret,
@@ -203,6 +216,54 @@ func TestParseStorageURLRejectsCredentials(t *testing.T) {
}
}
// TestParseStorageURLRejectsRcloneRemoteParams checks that an rclone
// remote name carrying parameters (rclone's "remote,param=value" form,
// which can hold a credential) is rejected without echoing the value,
// including the URLs that url.Parse itself refuses.
func TestParseStorageURLRejectsRcloneRemoteParams(t *testing.T) {
t.Parallel()
cases := []struct {
name string
raw string
secrets []string // must not appear in the error message
}{
{
name: "configured remote",
raw: "rclone://gdrive,client_secret=topsecret/backups",
secrets: []string{"topsecret"},
},
{
name: "remote defined inline",
raw: "rclone://:s3,access_key_id=KEY,secret_access_key=SECRET/bucket",
secrets: []string{"KEY", "SECRET"},
},
{
name: "colon in a parameter value",
raw: "rclone://gdrive,client_secret=ab:cd/backups",
secrets: []string{"cd"},
},
}
for _, tc := range cases {
t.Run(tc.name, func(t *testing.T) {
t.Parallel()
_, err := storage.ParseStorageURL(tc.raw)
if !errors.Is(err, storage.ErrRcloneRemoteParams) {
t.Fatalf("ParseStorageURL(%q) error = %v, want %v",
tc.raw, err, storage.ErrRcloneRemoteParams)
}
for _, secret := range tc.secrets {
if strings.Contains(err.Error(), secret) {
t.Errorf("error message leaked %q: %v", secret, err.Error())
}
}
})
}
}
// TestParseStorageURLParseFailureHidesURL checks that when url.Parse
// itself fails, the wrapped error carries only the inner cause, not the
// *url.Error whose text embeds the raw (possibly credential-bearing) URL.