Files
vaultik/internal/database/blobs.go
T
sneak 1073420e8b
check / check (pull_request) Successful in 2m52s
Trust only uploaded blobs for deduplication (closes #148)
An interrupted blob upload left the blob's chunks, blob_chunks, and
blobs rows committed before the upload was attempted, so a later run
deduplicated against data that never reached storage and produced a
snapshot that reported success but could not be restored.

Chosen fix (issue option b): a chunk counts as known only when a blob
holding it has uploaded_ts set, and each run drops un-uploaded blob
rows and the chunks they orphan at startup, so the affected data is
re-chunked and re-uploaded. Option a (withholding all row writes until
the upload succeeds) would entangle the packer's writes with upload
ordering; gated reads plus a startup repair is smaller and self-healing.

Blobs recorded with no remote backend are marked uploaded so the
invariant holds uniformly.

model: claude-opus-4-8
2026-09-21 23:22:29 +00:00

283 lines
6.8 KiB
Go

package database
import (
"context"
"database/sql"
"errors"
"fmt"
"time"
"sneak.berlin/go/vaultik/internal/log"
)
// BlobRepository provides access to the blobs table, which tracks the
// packed, encrypted storage units uploaded to the destination.
type BlobRepository struct {
db *DB
}
// NewBlobRepository creates a BlobRepository backed by db.
func NewBlobRepository(db *DB) *BlobRepository {
return &BlobRepository{db: db}
}
// Create inserts a blob row, using tx when non-nil.
func (r *BlobRepository) Create(ctx context.Context, tx *sql.Tx, blob *Blob) error {
query := `
INSERT INTO blobs (id, blob_hash, created_ts, finished_ts,
uncompressed_size, compressed_size, uploaded_ts)
VALUES (?, ?, ?, ?, ?, ?, ?)
`
var finishedTS, uploadedTS *int64
if blob.FinishedTS != nil {
ts := blob.FinishedTS.Unix()
finishedTS = &ts
}
if blob.UploadedTS != nil {
ts := blob.UploadedTS.Unix()
uploadedTS = &ts
}
var err error
if tx != nil {
_, err = tx.ExecContext(ctx, query, blob.ID, blob.Hash, blob.CreatedTS.Unix(),
finishedTS, blob.UncompressedSize, blob.CompressedSize, uploadedTS)
} else {
_, err = r.db.ExecWithLog(ctx, query, blob.ID, blob.Hash, blob.CreatedTS.Unix(),
finishedTS, blob.UncompressedSize, blob.CompressedSize, uploadedTS)
}
if err != nil {
return fmt.Errorf("inserting blob: %w", err)
}
return nil
}
// GetByHash returns the blob with the given content hash, or nil if no
// such blob exists.
func (r *BlobRepository) GetByHash(ctx context.Context, hash string) (*Blob, error) {
return r.getOne(ctx, "blob_hash", hash)
}
// GetByID retrieves a blob by its ID
func (r *BlobRepository) GetByID(ctx context.Context, id string) (*Blob, error) {
return r.getOne(ctx, "id", id)
}
// GetAll returns every blob row keyed by blob ID. Useful at restore
// start to translate the per-chunk blob_id references in chunkToBlobMap
// into blob hashes without doing one GetByID query per chunk.
func (r *BlobRepository) GetAll(ctx context.Context) (map[string]*Blob, error) {
query := `
SELECT id, blob_hash, created_ts, finished_ts,
uncompressed_size, compressed_size, uploaded_ts
FROM blobs
`
rows, err := r.db.conn.QueryContext(ctx, query)
if err != nil {
return nil, fmt.Errorf("querying blobs: %w", err)
}
defer func() {
err := rows.Close()
if err != nil {
Fatalf("failed to close rows: %v", err)
}
}()
out := make(map[string]*Blob)
for rows.Next() {
var (
blob Blob
createdTSUnix int64
finishedTSUnix, uploadedTSUnix sql.NullInt64
)
err := rows.Scan(
&blob.ID,
&blob.Hash,
&createdTSUnix,
&finishedTSUnix,
&blob.UncompressedSize,
&blob.CompressedSize,
&uploadedTSUnix,
)
if err != nil {
return nil, fmt.Errorf("scanning blob: %w", err)
}
blob.CreatedTS = time.Unix(createdTSUnix, 0).UTC()
if finishedTSUnix.Valid {
ts := time.Unix(finishedTSUnix.Int64, 0).UTC()
blob.FinishedTS = &ts
}
if uploadedTSUnix.Valid {
ts := time.Unix(uploadedTSUnix.Int64, 0).UTC()
blob.UploadedTS = &ts
}
out[blob.ID.String()] = &blob
}
return out, rows.Err()
}
// UpdateFinished updates a blob when it's finalized
func (r *BlobRepository) UpdateFinished(
ctx context.Context,
tx *sql.Tx,
id string,
hash string,
uncompressedSize, compressedSize int64,
) error {
query := `
UPDATE blobs
SET blob_hash = ?, finished_ts = ?, uncompressed_size = ?, compressed_size = ?
WHERE id = ?
`
now := time.Now().UTC().Unix()
var err error
if tx != nil {
_, err = tx.ExecContext(ctx, query, hash, now, uncompressedSize, compressedSize, id)
} else {
_, err = r.db.ExecWithLog(ctx, query, hash, now, uncompressedSize, compressedSize, id)
}
if err != nil {
return fmt.Errorf("updating blob: %w", err)
}
return nil
}
// UpdateUploaded marks a blob as uploaded
func (r *BlobRepository) UpdateUploaded(
ctx context.Context, tx *sql.Tx, id string,
) error {
query := `
UPDATE blobs
SET uploaded_ts = ?
WHERE id = ?
`
now := time.Now().UTC().Unix()
var err error
if tx != nil {
_, err = tx.ExecContext(ctx, query, now, id)
} else {
_, err = r.db.ExecWithLog(ctx, query, now, id)
}
if err != nil {
return fmt.Errorf("marking blob as uploaded: %w", err)
}
return nil
}
// DeleteOrphaned deletes blobs that are not referenced by any snapshot
func (r *BlobRepository) DeleteOrphaned(ctx context.Context) error {
query := `
DELETE FROM blobs
WHERE NOT EXISTS (
SELECT 1 FROM snapshot_blobs
WHERE snapshot_blobs.blob_id = blobs.id
)
`
result, err := r.db.ExecWithLog(ctx, query)
if err != nil {
return fmt.Errorf("deleting orphaned blobs: %w", err)
}
rowsAffected, _ := result.RowsAffected()
if rowsAffected > 0 {
log.Debug("Deleted orphaned blobs", "count", rowsAffected)
}
return nil
}
// DeleteUnuploaded deletes blob rows whose upload never completed
// (uploaded_ts IS NULL) and returns how many were removed. Their
// blob_chunks rows are removed by the ON DELETE CASCADE foreign key.
// A blob is only ever attached to a snapshot once its upload has been
// recorded, so an un-uploaded blob is never referenced by a completed
// snapshot: dropping it discards chunk rows that point at data which
// was never stored remotely, so the affected content is re-chunked and
// re-uploaded on the next run.
func (r *BlobRepository) DeleteUnuploaded(ctx context.Context) (int64, error) {
query := `DELETE FROM blobs WHERE uploaded_ts IS NULL`
result, err := r.db.ExecWithLog(ctx, query)
if err != nil {
return 0, fmt.Errorf("deleting un-uploaded blobs: %w", err)
}
rowsAffected, _ := result.RowsAffected()
if rowsAffected > 0 {
log.Debug("Deleted un-uploaded blobs", "count", rowsAffected)
}
return rowsAffected, nil
}
// getOne fetches a single blob row matched on the given column, or
// (nil, nil) when no row matches.
func (r *BlobRepository) getOne(
ctx context.Context, column, value string,
) (*Blob, error) {
query := `
SELECT id, blob_hash, created_ts, finished_ts,
uncompressed_size, compressed_size, uploaded_ts
FROM blobs
WHERE ` + column + ` = ?`
var (
blob Blob
createdTSUnix int64
finishedTSUnix, uploadedTSUnix sql.NullInt64
)
err := r.db.conn.QueryRowContext(ctx, query, value).Scan(
&blob.ID,
&blob.Hash,
&createdTSUnix,
&finishedTSUnix,
&blob.UncompressedSize,
&blob.CompressedSize,
&uploadedTSUnix,
)
if errors.Is(err, sql.ErrNoRows) {
return nil, nil //nolint:nilnil // nil,nil signals not-found; callers check nil
}
if err != nil {
return nil, fmt.Errorf("querying blob: %w", err)
}
blob.CreatedTS = time.Unix(createdTSUnix, 0).UTC()
if finishedTSUnix.Valid {
ts := time.Unix(finishedTSUnix.Int64, 0).UTC()
blob.FinishedTS = &ts
}
if uploadedTSUnix.Valid {
ts := time.Unix(uploadedTSUnix.Int64, 0).UTC()
blob.UploadedTS = &ts
}
return &blob, nil
}