Files
seaweedfs/weed/filer/filechunk_manifest_rollback_test.go
T
Eliah RusinandChris Lu cfa8afec92 filer: guard FoundationDB 100KB value limit and pack earlier (#11161)
* filer: guard the FoundationDB value size limit, not the transaction limit

An entry's whole chunk list is one FoundationDB value, and FDB caps a value at
100,000 bytes while a transaction may reach 10MB. UpdateEntry checked the
transaction limit, so every entry between the two limits passed the guard and
was rejected by FDB itself with error 2103 (Value length exceeds limit). The
failure surfaced inside the store rather than at the guard, so the S3 layer
dropped the connection and clients saw a network fault instead of an error.

Check the value limit in UpdateEntry and KvPut instead, after gzip and before
the transaction, with an error that names the limit it hit. The removed
transaction-size constant guarded nothing else: DeleteFolderChildren batches by
entry count.

Refs #11158

* filer: fold at 500 chunks in the foundationdb build

Manifest packing is what keeps a large file's entry small, but it only ran once
a flat chunk list reached 10000 chunks. A FoundationDB value stops at 100,000
bytes and an entry's whole chunk list is one value, which at ~100 bytes per
chunk record is about 1000 chunks -- so on FDB the write always failed before
packing could help: a 3.3 GiB PutObject at the default -maxMB=4 was already
past the limit.

FoundationDB support is its own build (`go build -tags foundationdb`, shipped
as its own image), so the batch is a build-time choice and needs no negotiation
at run time. The tagged build folds at 500, every other build keeps 10000 and
is untouched.

500 is not arbitrary: a single fold level leaves (chunks/batch) manifest
pointers plus up to (batch-1) unfolded chunks in the entry, so the reachable
chunk count is highest when the two terms are near equal. For a 100,000-byte
budget that optimum is 500, which holds an entry inside the limit up to
~250,000 chunks -- ~1 TB at -maxMB=4, against ~4 GB before. Larger files need
nested packing, which no batch size substitutes for.

One binary serves every role in that image, so the filer and each client that
folds -- S3, mount, WebDAV, weed shell, filer.copy -- agree on the batch by
construction. A binary built with the tag but pointed at another store folds
earlier than that store requires, costing one manifest blob per 500 chunks and
one read to resolve it.

Fixes #11158

* filer: fold with rollback inside MaybeManifestize, not beside it

A fold that fails midway has already uploaded manifest blobs for its earlier
batches, and returns only the data chunks -- dropping the manifests it had
separated out of the caller's list. Both were wrong in ways that mattered:

  - AppendToEntry assigned that shortened list straight to entry.Chunks and
    created the entry, so an append to an already-folded file whose fold
    failed lost every previously folded chunk. weed mount had the same shape.
  - cleanupChunks logged the error as "not good, but should be ok" and then
    returned it through a named result, failing the whole CreateEntry or
    UpdateEntry, while the blobs it had written stayed behind referenced by
    nothing.

The S3 path was alone in handling this, through a private helper beside
MaybeManifestize. A second entry point next to the one everything else calls
just means the wrong one gets used, so the behaviour moves inside
MaybeManifestize: on failure it returns inputChunks as it received them, and
hands the blobs it saved to a deleteChunks callback. The filer, S3 and
filer.copy pass their existing deleters -- filer.copy already cleans up this
way after a failed upload -- and mount, WebDAV and weed shell pass nil, which
reports the blobs rather than collecting them, as before. Each caller keeps its
own error policy: the filer HTTP PUT path and filer.copy still fail the request,
the rest still continue with the flat list, which is a correct entry.

---------

Co-authored-by: Chris Lu <chris.lu@gmail.com>
2026-09-04 23:12:27 -07:00

148 lines
4.8 KiB
Go

package filer
import (
"fmt"
"io"
"testing"
"github.com/seaweedfs/seaweedfs/weed/pb/filer_pb"
)
func flatTestChunks(n int) []*filer_pb.FileChunk {
chunks := make([]*filer_pb.FileChunk, n)
for i := range chunks {
chunks[i] = &filer_pb.FileChunk{
FileId: fmt.Sprintf("1,%x", i+1),
Offset: int64(i) * 8,
Size: 8,
}
}
return chunks
}
type fakeManifestStore struct {
saves int
failOn int // 1-based save call to fail on; 0 = never
deleted []*filer_pb.FileChunk
}
func (s *fakeManifestStore) save(reader io.Reader, name string, offset int64, tsNs int64, expectedDataSize uint64) (*filer_pb.FileChunk, error) {
s.saves++
if s.saves == s.failOn {
return nil, fmt.Errorf("save %d failed", s.saves)
}
if _, err := io.Copy(io.Discard, reader); err != nil {
return nil, err
}
return &filer_pb.FileChunk{FileId: fmt.Sprintf("2,%x", s.saves), Offset: offset}, nil
}
func (s *fakeManifestStore) delete(chunks []*filer_pb.FileChunk) {
s.deleted = append(s.deleted, chunks...)
}
func TestMaybeManifestizeFolds(t *testing.T) {
store := &fakeManifestStore{}
chunks := flatTestChunks(ManifestBatch + 50)
result, err := MaybeManifestize(store.save, store.delete, chunks)
if err != nil {
t.Fatalf("MaybeManifestize: %v", err)
}
manifests, data := SeparateManifestChunks(result)
if len(manifests) != 1 || len(data) != 50 {
t.Fatalf("expected 1 manifest + 50 flat chunks, got %d + %d", len(manifests), len(data))
}
if store.saves != 1 || len(store.deleted) != 0 {
t.Errorf("expected 1 save and no deletes, got %d saves, %d deleted", store.saves, len(store.deleted))
}
if manifests[0].Offset != 0 || manifests[0].Size != uint64(ManifestBatch*8) {
t.Errorf("manifest span [%d,%d) wrong", manifests[0].Offset, manifests[0].Size)
}
}
func TestMaybeManifestizeBelowThreshold(t *testing.T) {
store := &fakeManifestStore{}
chunks := flatTestChunks(ManifestBatch - 1)
result, err := MaybeManifestize(store.save, store.delete, chunks)
if err != nil {
t.Fatalf("MaybeManifestize: %v", err)
}
if len(result) != len(chunks) || store.saves != 0 {
t.Fatalf("expected untouched flat list, got %d chunks, %d saves", len(result), store.saves)
}
}
func TestMaybeManifestizeSkipsSse(t *testing.T) {
store := &fakeManifestStore{}
chunks := flatTestChunks(ManifestBatch + 50)
chunks[0].SseType = filer_pb.SSEType_SSE_S3
result, err := MaybeManifestize(store.save, store.delete, chunks)
if err != nil {
t.Fatalf("MaybeManifestize: %v", err)
}
if len(result) != len(chunks) || store.saves != 0 {
t.Fatalf("SSE chunks must not be folded, got %d chunks, %d saves", len(result), store.saves)
}
}
// A fold that fails midway must hand back the caller's own list and delete the
// manifest blobs its earlier batches already uploaded.
func TestMaybeManifestizeRollsBackPartialFold(t *testing.T) {
store := &fakeManifestStore{failOn: 2}
chunks := flatTestChunks(2*ManifestBatch + 50)
result, err := MaybeManifestize(store.save, store.delete, chunks)
if err == nil {
t.Fatal("expected the failed save to be reported")
}
if len(result) != len(chunks) {
t.Fatalf("expected fallback to %d flat chunks, got %d", len(chunks), len(result))
}
if HasChunkManifest(result) {
t.Error("fallback list must not contain manifest chunks")
}
if len(store.deleted) != 1 || store.deleted[0].FileId != "2,1" {
t.Fatalf("expected the first saved blob deleted, got %+v", store.deleted)
}
}
// The manifests a caller came in with are not re-folded, and must survive a
// fold that fails on the flat remainder: returning only the data chunks would
// lose everything they cover.
func TestMaybeManifestizeKeepsExistingManifestsOnFailure(t *testing.T) {
store := &fakeManifestStore{failOn: 2}
existing := &filer_pb.FileChunk{FileId: "m,1", IsChunkManifest: true, Offset: 0, Size: 24}
chunks := append([]*filer_pb.FileChunk{existing}, flatTestChunks(2*ManifestBatch+50)...)
result, _ := MaybeManifestize(store.save, store.delete, chunks)
if len(result) != len(chunks) {
t.Fatalf("expected original %d chunks, got %d", len(chunks), len(result))
}
if result[0].FileId != "m,1" || !result[0].IsChunkManifest {
t.Fatalf("existing manifest dropped: %+v", result[0])
}
if len(store.deleted) != 1 || store.deleted[0].FileId != "2,1" {
t.Fatalf("expected the first saved blob deleted, got %+v", store.deleted)
}
}
// Without a deleter the fold still falls back to the caller's list; the blobs
// it saved are only reported.
func TestMaybeManifestizeRollbackWithoutDeleter(t *testing.T) {
store := &fakeManifestStore{failOn: 2}
chunks := flatTestChunks(2*ManifestBatch + 50)
result, err := MaybeManifestize(store.save, nil, chunks)
if err == nil {
t.Fatal("expected the failed save to be reported")
}
if len(result) != len(chunks) || HasChunkManifest(result) {
t.Fatalf("expected fallback to %d flat chunks, got %d", len(chunks), len(result))
}
}