storj/storagenode/piecestore/endpoint_test.go
Isaac Hess a785d37157 storagenode/pieces: Process deletes asynchronously
To improve delete performance, we want to process deletes asynchronously
once the message has been received from the satellite. This change makes
it so that storagenodes will send the delete request to a piece Deleter,
which will process a "best-effort" delete asynchronously and return a
success message to the satellite.

There is a configurable number of max delete workers and a max delete
queue size.

Change-Id: I016b68031f9065a9b09224f161b6783e18cf21e5
2020-04-23 11:51:19 -06:00

661 lines
20 KiB
Go

// Copyright (C) 2019 Storj Labs, Inc.
// See LICENSE for copying information
package piecestore_test
import (
"io"
"strings"
"sync/atomic"
"testing"
"time"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
"github.com/zeebo/errs"
"go.uber.org/zap"
"go.uber.org/zap/zaptest"
"golang.org/x/sync/errgroup"
"storj.io/common/errs2"
"storj.io/common/memory"
"storj.io/common/pb"
"storj.io/common/pkcrypto"
"storj.io/common/rpc/rpcstatus"
"storj.io/common/signing"
"storj.io/common/storj"
"storj.io/common/testcontext"
"storj.io/common/testrand"
"storj.io/storj/private/testplanet"
"storj.io/storj/storagenode"
"storj.io/storj/storagenode/bandwidth"
"storj.io/uplink/private/piecestore"
)
func TestUploadAndPartialDownload(t *testing.T) {
testplanet.Run(t, testplanet.Config{
SatelliteCount: 1, StorageNodeCount: 6, UplinkCount: 1,
}, func(t *testing.T, ctx *testcontext.Context, planet *testplanet.Planet) {
expectedData := testrand.Bytes(100 * memory.KiB)
err := planet.Uplinks[0].Upload(ctx, planet.Satellites[0], "testbucket", "test/path", expectedData)
assert.NoError(t, err)
var totalDownload int64
for _, tt := range []struct {
offset, size int64
}{
{0, 1510},
{1513, 1584},
{13581, 4783},
} {
if piecestore.DefaultConfig.InitialStep < tt.size {
t.Fatal("test expects initial step to be larger than size to download")
}
totalDownload += piecestore.DefaultConfig.InitialStep
download, cleanup, err := planet.Uplinks[0].DownloadStreamRange(ctx, planet.Satellites[0], "testbucket", "test/path", tt.offset, -1)
require.NoError(t, err)
defer ctx.Check(cleanup)
data := make([]byte, tt.size)
n, err := io.ReadFull(download, data)
require.NoError(t, err)
assert.Equal(t, int(tt.size), n)
assert.Equal(t, expectedData[tt.offset:tt.offset+tt.size], data)
require.NoError(t, download.Close())
}
var totalBandwidthUsage bandwidth.Usage
for _, storagenode := range planet.StorageNodes {
usage, err := storagenode.DB.Bandwidth().Summary(ctx, time.Now().Add(-10*time.Hour), time.Now().Add(10*time.Hour))
require.NoError(t, err)
totalBandwidthUsage.Add(usage)
}
err = planet.Uplinks[0].DeleteObject(ctx, planet.Satellites[0], "testbucket", "test/path")
require.NoError(t, err)
_, err = planet.Uplinks[0].Download(ctx, planet.Satellites[0], "testbucket", "test/path")
require.Error(t, err)
// check rough limits for the upload and download
totalUpload := int64(len(expectedData))
t.Log(totalUpload, totalBandwidthUsage.Put, int64(len(planet.StorageNodes))*totalUpload)
assert.True(t, totalUpload < totalBandwidthUsage.Put && totalBandwidthUsage.Put < int64(len(planet.StorageNodes))*totalUpload)
t.Log(totalDownload, totalBandwidthUsage.Get, int64(len(planet.StorageNodes))*totalDownload)
assert.True(t, totalBandwidthUsage.Get < int64(len(planet.StorageNodes))*totalDownload)
})
}
func TestUpload(t *testing.T) {
testplanet.Run(t, testplanet.Config{
SatelliteCount: 1, StorageNodeCount: 1, UplinkCount: 1,
}, func(t *testing.T, ctx *testcontext.Context, planet *testplanet.Planet) {
client, err := planet.Uplinks[0].DialPiecestore(ctx, planet.StorageNodes[0])
require.NoError(t, err)
defer ctx.Check(client.Close)
for _, tt := range []struct {
pieceID storj.PieceID
contentLength memory.Size
action pb.PieceAction
err string
}{
{ // should successfully store data
pieceID: storj.PieceID{1},
contentLength: 50 * memory.KiB,
action: pb.PieceAction_PUT,
err: "",
},
{ // should err with piece ID not specified
pieceID: storj.PieceID{},
contentLength: 1 * memory.KiB,
action: pb.PieceAction_PUT,
err: "missing piece id",
},
{ // should err because invalid action
pieceID: storj.PieceID{1},
contentLength: 1 * memory.KiB,
action: pb.PieceAction_GET,
err: "expected put or put repair action got GET",
},
} {
data := testrand.Bytes(tt.contentLength)
serialNumber := testrand.SerialNumber()
orderLimit, piecePrivateKey := GenerateOrderLimit(
t,
planet.Satellites[0].ID(),
planet.StorageNodes[0].ID(),
tt.pieceID,
tt.action,
serialNumber,
24*time.Hour,
24*time.Hour,
int64(len(data)),
)
signer := signing.SignerFromFullIdentity(planet.Satellites[0].Identity)
orderLimit, err = signing.SignOrderLimit(ctx, signer, orderLimit)
require.NoError(t, err)
uploader, err := client.Upload(ctx, orderLimit, piecePrivateKey)
require.NoError(t, err)
_, err = uploader.Write(data)
require.NoError(t, err)
pieceHash, err := uploader.Commit(ctx)
if tt.err != "" {
require.Error(t, err)
require.Contains(t, err.Error(), tt.err)
} else {
require.NoError(t, err)
expectedHash := pkcrypto.SHA256Hash(data)
assert.Equal(t, expectedHash, pieceHash.Hash)
signee := signing.SignerFromFullIdentity(planet.StorageNodes[0].Identity)
require.NoError(t, signing.VerifyPieceHashSignature(ctx, signee, pieceHash))
}
}
})
}
func TestDownload(t *testing.T) {
testplanet.Run(t, testplanet.Config{
SatelliteCount: 1, StorageNodeCount: 1, UplinkCount: 1,
}, func(t *testing.T, ctx *testcontext.Context, planet *testplanet.Planet) {
pieceID := storj.PieceID{1}
expectedData, _, _ := uploadPiece(t, ctx, pieceID, planet.StorageNodes[0], planet.Uplinks[0], planet.Satellites[0])
client, err := planet.Uplinks[0].DialPiecestore(ctx, planet.StorageNodes[0])
require.NoError(t, err)
for _, tt := range []struct {
pieceID storj.PieceID
action pb.PieceAction
errs []string
}{
{ // should successfully download data
pieceID: pieceID,
action: pb.PieceAction_GET,
},
{ // should err with piece ID not specified
pieceID: storj.PieceID{},
action: pb.PieceAction_GET,
errs: []string{"missing piece id"},
},
{ // should err with piece ID not specified
pieceID: storj.PieceID{2},
action: pb.PieceAction_GET,
errs: []string{"file does not exist", "The system cannot find the path specified"},
},
{ // should err with invalid action
pieceID: pieceID,
action: pb.PieceAction_PUT,
errs: []string{"expected get or get repair or audit action got PUT"},
},
} {
serialNumber := testrand.SerialNumber()
orderLimit, piecePrivateKey := GenerateOrderLimit(
t,
planet.Satellites[0].ID(),
planet.StorageNodes[0].ID(),
tt.pieceID,
tt.action,
serialNumber,
24*time.Hour,
24*time.Hour,
int64(len(expectedData)),
)
signer := signing.SignerFromFullIdentity(planet.Satellites[0].Identity)
orderLimit, err = signing.SignOrderLimit(ctx, signer, orderLimit)
require.NoError(t, err)
downloader, err := client.Download(ctx, orderLimit, piecePrivateKey, 0, int64(len(expectedData)))
require.NoError(t, err)
buffer := make([]byte, len(expectedData))
n, err := downloader.Read(buffer)
if len(tt.errs) > 0 {
} else {
require.NoError(t, err)
require.Equal(t, expectedData, buffer[:n])
}
err = downloader.Close()
if len(tt.errs) > 0 {
require.Error(t, err)
require.True(t, strings.Contains(err.Error(), tt.errs[0]) || strings.Contains(err.Error(), tt.errs[1]), err.Error())
} else {
require.NoError(t, err)
}
// these should only be not-nil if action = pb.PieceAction_GET_REPAIR
hash, originalLimit := downloader.GetHashAndLimit()
require.Nil(t, hash)
require.Nil(t, originalLimit)
}
})
}
func TestDownloadGetRepair(t *testing.T) {
testplanet.Run(t, testplanet.Config{
SatelliteCount: 1, StorageNodeCount: 1, UplinkCount: 1,
}, func(t *testing.T, ctx *testcontext.Context, planet *testplanet.Planet) {
pieceID := storj.PieceID{1}
expectedData, ulOrderLimit, originHash := uploadPiece(
t, ctx, pieceID, planet.StorageNodes[0], planet.Uplinks[0], planet.Satellites[0],
)
client, err := planet.Uplinks[0].DialPiecestore(ctx, planet.StorageNodes[0])
require.NoError(t, err)
serialNumber := testrand.SerialNumber()
dlOrderLimit, piecePrivateKey := GenerateOrderLimit(
t,
planet.Satellites[0].ID(),
planet.StorageNodes[0].ID(),
storj.PieceID{1},
pb.PieceAction_GET_REPAIR,
serialNumber,
24*time.Hour,
24*time.Hour,
int64(len(expectedData)),
)
signer := signing.SignerFromFullIdentity(planet.Satellites[0].Identity)
dlOrderLimit, err = signing.SignOrderLimit(ctx, signer, dlOrderLimit)
require.NoError(t, err)
downloader, err := client.Download(ctx, dlOrderLimit, piecePrivateKey, 0, int64(len(expectedData)))
require.NoError(t, err)
buffer := make([]byte, len(expectedData))
n, err := downloader.Read(buffer)
require.NoError(t, err)
require.Equal(t, expectedData, buffer[:n])
err = downloader.Close()
require.NoError(t, err)
hash, originLimit := downloader.GetHashAndLimit()
require.NotNil(t, hash)
require.Equal(t, originHash.Hash, hash.Hash)
require.Equal(t, originHash.PieceId, hash.PieceId)
require.NotNil(t, originLimit)
require.Equal(t, originLimit.Action, ulOrderLimit.Action)
require.Equal(t, originLimit.Limit, ulOrderLimit.Limit)
require.Equal(t, originLimit.PieceId, ulOrderLimit.PieceId)
require.Equal(t, originLimit.SatelliteId, ulOrderLimit.SatelliteId)
require.Equal(t, originLimit.SerialNumber, ulOrderLimit.SerialNumber)
require.Equal(t, originLimit.SatelliteSignature, ulOrderLimit.SatelliteSignature)
require.Equal(t, originLimit.UplinkPublicKey, ulOrderLimit.UplinkPublicKey)
})
}
func TestDelete(t *testing.T) {
testplanet.Run(t, testplanet.Config{
SatelliteCount: 1, StorageNodeCount: 1, UplinkCount: 1,
}, func(t *testing.T, ctx *testcontext.Context, planet *testplanet.Planet) {
pieceID := storj.PieceID{1}
uploadPiece(t, ctx, pieceID, planet.StorageNodes[0], planet.Uplinks[0], planet.Satellites[0])
client, err := planet.Uplinks[0].DialPiecestore(ctx, planet.StorageNodes[0])
require.NoError(t, err)
for _, tt := range []struct {
pieceID storj.PieceID
action pb.PieceAction
err string
}{
{ // should successfully delete data
pieceID: pieceID,
action: pb.PieceAction_DELETE,
err: "",
},
{ // should err with piece ID not found
pieceID: storj.PieceID{99},
action: pb.PieceAction_DELETE,
err: "", // TODO should this return error
},
{ // should err with piece ID not specified
pieceID: storj.PieceID{},
action: pb.PieceAction_DELETE,
err: "missing piece id",
},
{ // should err due to incorrect action
pieceID: pieceID,
action: pb.PieceAction_GET,
err: "expected delete action got GET",
},
} {
serialNumber := testrand.SerialNumber()
orderLimit, piecePrivateKey := GenerateOrderLimit(
t,
planet.Satellites[0].ID(),
planet.StorageNodes[0].ID(),
tt.pieceID,
tt.action,
serialNumber,
24*time.Hour,
24*time.Hour,
100,
)
signer := signing.SignerFromFullIdentity(planet.Satellites[0].Identity)
orderLimit, err = signing.SignOrderLimit(ctx, signer, orderLimit)
require.NoError(t, err)
err := client.Delete(ctx, orderLimit, piecePrivateKey)
if tt.err != "" {
require.Error(t, err)
require.Contains(t, err.Error(), tt.err)
} else {
require.NoError(t, err)
}
}
})
}
func TestDeletePieces(t *testing.T) {
testplanet.Run(t, testplanet.Config{
SatelliteCount: 1, StorageNodeCount: 1, UplinkCount: 1,
}, func(t *testing.T, ctx *testcontext.Context, planet *testplanet.Planet) {
var (
planetSat = planet.Satellites[0]
planetSN = planet.StorageNodes[0]
)
var client *piecestore.Client
{
dossier, err := planetSat.Overlay.DB.Get(ctx.Context, planetSN.ID())
require.NoError(t, err)
client, err = piecestore.Dial(
ctx.Context, planetSat.Dialer, &dossier.Node, zaptest.NewLogger(t), piecestore.Config{},
)
require.NoError(t, err)
}
t.Run("Ok", func(t *testing.T) {
pieceIDs := []storj.PieceID{testrand.PieceID(), testrand.PieceID(), testrand.PieceID(), testrand.PieceID()}
dataArray := make([][]byte, len(pieceIDs))
for i, pieceID := range pieceIDs {
dataArray[i], _, _ = uploadPiece(t, ctx, pieceID, planetSN, planet.Uplinks[0], planetSat)
}
err := client.DeletePieces(ctx.Context, pieceIDs...)
require.NoError(t, err)
for i, pieceID := range pieceIDs {
_, err = downloadPiece(t, ctx, pieceID, int64(len(dataArray[i])), planetSN, planet.Uplinks[0], planetSat)
require.Error(t, err)
}
require.Condition(t, func() bool {
return strings.Contains(err.Error(), "file does not exist") ||
strings.Contains(err.Error(), "The system cannot find the path specified")
}, "unexpected error message")
})
t.Run("Ok: one piece to delete is missing", func(t *testing.T) {
missingPieceID := testrand.PieceID()
pieceIDs := []storj.PieceID{testrand.PieceID(), testrand.PieceID(), testrand.PieceID(), testrand.PieceID()}
dataArray := make([][]byte, len(pieceIDs))
for i, pieceID := range pieceIDs {
dataArray[i], _, _ = uploadPiece(t, ctx, pieceID, planetSN, planet.Uplinks[0], planetSat)
}
err := client.DeletePieces(ctx.Context, append(pieceIDs, missingPieceID)...)
require.NoError(t, err)
for i, pieceID := range pieceIDs {
_, err = downloadPiece(t, ctx, pieceID, int64(len(dataArray[i])), planetSN, planet.Uplinks[0], planetSat)
require.Error(t, err)
}
require.Condition(t, func() bool {
return strings.Contains(err.Error(), "file does not exist") ||
strings.Contains(err.Error(), "The system cannot find the path specified")
}, "unexpected error message")
})
t.Run("Ok: no piece deleted", func(t *testing.T) {
pieceID := testrand.PieceID()
data, _, _ := uploadPiece(t, ctx, pieceID, planetSN, planet.Uplinks[0], planetSat)
err := client.DeletePieces(ctx.Context)
require.NoError(t, err)
downloaded, err := downloadPiece(t, ctx, pieceID, int64(len(data)), planetSN, planet.Uplinks[0], planetSat)
require.NoError(t, err)
require.Equal(t, data, downloaded)
})
t.Run("error: permission denied", func(t *testing.T) {
pieceID := testrand.PieceID()
data, _, _ := uploadPiece(t, ctx, pieceID, planetSN, planet.Uplinks[0], planetSat)
client, err := planet.Uplinks[0].DialPiecestore(ctx, planetSN)
require.NoError(t, err)
err = client.DeletePieces(ctx.Context, pieceID)
require.Error(t, err)
require.Equal(t, rpcstatus.PermissionDenied, rpcstatus.Code(err))
downloaded, err := downloadPiece(t, ctx, pieceID, int64(len(data)), planetSN, planet.Uplinks[0], planetSat)
require.NoError(t, err)
require.Equal(t, data, downloaded)
})
})
}
func TestTooManyRequests(t *testing.T) {
t.Skip("flaky, because of EOF issues")
const uplinkCount = 6
const maxConcurrent = 3
const expectedFailures = uplinkCount - maxConcurrent
testplanet.Run(t, testplanet.Config{
SatelliteCount: 1, StorageNodeCount: 1, UplinkCount: uplinkCount,
Reconfigure: testplanet.Reconfigure{
StorageNode: func(index int, config *storagenode.Config) {
config.Storage2.MaxConcurrentRequests = maxConcurrent
},
},
}, func(t *testing.T, ctx *testcontext.Context, planet *testplanet.Planet) {
doneWaiting := make(chan struct{})
failedCount := int64(expectedFailures)
uploads, _ := errgroup.WithContext(ctx)
defer ctx.Check(uploads.Wait)
for i, uplink := range planet.Uplinks {
i, uplink := i, uplink
uploads.Go(func() (err error) {
storageNode := planet.StorageNodes[0].Local()
config := piecestore.DefaultConfig
config.UploadBufferSize = 0 // disable buffering so we can detect write error early
client, err := piecestore.Dial(ctx, uplink.Dialer, &storageNode.Node, uplink.Log, config)
if err != nil {
return err
}
defer func() {
if cerr := client.Close(); cerr != nil {
uplink.Log.Error("close failed", zap.Error(cerr))
err = errs.Combine(err, cerr)
}
}()
pieceID := storj.PieceID{byte(i + 1)}
serialNumber := testrand.SerialNumber()
orderLimit, piecePrivateKey := GenerateOrderLimit(
t,
planet.Satellites[0].ID(),
planet.StorageNodes[0].ID(),
pieceID,
pb.PieceAction_PUT,
serialNumber,
24*time.Hour,
24*time.Hour,
int64(10000),
)
satelliteSigner := signing.SignerFromFullIdentity(planet.Satellites[0].Identity)
orderLimit, err = signing.SignOrderLimit(ctx, satelliteSigner, orderLimit)
if err != nil {
return err
}
upload, err := client.Upload(ctx, orderLimit, piecePrivateKey)
if err != nil {
if errs2.IsRPC(err, rpcstatus.Unavailable) {
if atomic.AddInt64(&failedCount, -1) == 0 {
close(doneWaiting)
}
return nil
}
uplink.Log.Error("upload failed", zap.Stringer("Piece ID", pieceID), zap.Error(err))
return err
}
_, err = upload.Write(make([]byte, orderLimit.Limit))
if err != nil {
if errs2.IsRPC(err, rpcstatus.Unavailable) {
if atomic.AddInt64(&failedCount, -1) == 0 {
close(doneWaiting)
}
return nil
}
uplink.Log.Error("write failed", zap.Stringer("Piece ID", pieceID), zap.Error(err))
return err
}
_, err = upload.Commit(ctx)
if err != nil {
if errs2.IsRPC(err, rpcstatus.Unavailable) {
if atomic.AddInt64(&failedCount, -1) == 0 {
close(doneWaiting)
}
return nil
}
uplink.Log.Error("commit failed", zap.Stringer("Piece ID", pieceID), zap.Error(err))
return err
}
return nil
})
}
})
}
func GenerateOrderLimit(t *testing.T, satellite storj.NodeID, storageNode storj.NodeID, pieceID storj.PieceID, action pb.PieceAction, serialNumber storj.SerialNumber, pieceExpiration, orderExpiration time.Duration, limit int64) (*pb.OrderLimit, storj.PiecePrivateKey) {
piecePublicKey, piecePrivateKey, err := storj.NewPieceKey()
require.NoError(t, err)
now := time.Now()
return &pb.OrderLimit{
SatelliteId: satellite,
UplinkPublicKey: piecePublicKey,
StorageNodeId: storageNode,
PieceId: pieceID,
Action: action,
SerialNumber: serialNumber,
OrderCreation: time.Now(),
OrderExpiration: now.Add(orderExpiration),
PieceExpiration: now.Add(pieceExpiration),
Limit: limit,
}, piecePrivateKey
}
// uploadPiece uploads piece to storageNode.
func uploadPiece(
t *testing.T, ctx *testcontext.Context, piece storj.PieceID, storageNode *testplanet.StorageNode,
uplink *testplanet.Uplink, satellite *testplanet.Satellite,
) (uploadedData []byte, _ *pb.OrderLimit, _ *pb.PieceHash) {
t.Helper()
client, err := uplink.DialPiecestore(ctx, storageNode)
require.NoError(t, err)
defer ctx.Check(client.Close)
serialNumber := testrand.SerialNumber()
uploadedData = testrand.Bytes(10 * memory.KiB)
orderLimit, piecePrivateKey := GenerateOrderLimit(
t,
satellite.ID(),
storageNode.ID(),
piece,
pb.PieceAction_PUT,
serialNumber,
24*time.Hour,
24*time.Hour,
int64(len(uploadedData)),
)
signer := signing.SignerFromFullIdentity(satellite.Identity)
orderLimit, err = signing.SignOrderLimit(ctx, signer, orderLimit)
require.NoError(t, err)
uploader, err := client.Upload(ctx, orderLimit, piecePrivateKey)
require.NoError(t, err)
_, err = uploader.Write(uploadedData)
require.NoError(t, err)
hash, err := uploader.Commit(ctx)
require.NoError(t, err)
return uploadedData, orderLimit, hash
}
// downloadPiece downlodads piece from storageNode.
func downloadPiece(
t *testing.T, ctx *testcontext.Context, piece storj.PieceID, limit int64,
storageNode *testplanet.StorageNode, uplink *testplanet.Uplink, satellite *testplanet.Satellite,
) (pieceData []byte, err error) {
t.Helper()
serialNumber := testrand.SerialNumber()
orderLimit, piecePrivateKey := GenerateOrderLimit(
t,
satellite.ID(),
storageNode.ID(),
piece,
pb.PieceAction_GET,
serialNumber,
24*time.Hour,
24*time.Hour,
limit,
)
signer := signing.SignerFromFullIdentity(satellite.Identity)
orderLimit, err = signing.SignOrderLimit(ctx.Context, signer, orderLimit)
require.NoError(t, err)
client, err := uplink.DialPiecestore(ctx, storageNode)
require.NoError(t, err)
downloader, err := client.Download(ctx.Context, orderLimit, piecePrivateKey, 0, limit)
require.NoError(t, err)
defer func() {
if err != nil {
// Ignore err in Close if an error happened in Download because it's also
// returned by Close.
_ = downloader.Close()
return
}
err = downloader.Close()
}()
buffer := make([]byte, limit)
n, err := downloader.Read(buffer)
return buffer[:n], err
}