storj/satellite/metabase/rangedloop/service_test.go
paul cannon ae5947327b satellite/accounting: Use metabase.AliasPiece with tally observer
We want to eliminate usages of LoopSegmentEntry.Pieces, because it is
costing a lot of cpu time to look up node IDs with every piece of every
segment we read.

In this change, we are eliminating use of LoopSegmentEntry.Pieces in the
node tally observer (both the ranged loop and segments loop variants).
It is not necessary to have a fully resolved nodeID until it is time to
store totals in the database. We can use NodeAliases as the map key
instead, and resolve NodeIDs just before storing totals.

Refs: https://github.com/storj/storj/issues/5622
Change-Id: Iec12aa393072436d7c22cc5a4ae1b63966cbcc18
2023-03-29 12:24:05 +00:00

512 lines
14 KiB
Go

// Copyright (C) 2022 Storj Labs, Inc.
// See LICENSE for copying information.
package rangedloop_test
import (
"context"
"errors"
"fmt"
"sort"
"strconv"
"sync"
"sync/atomic"
"testing"
"time"
"github.com/stretchr/testify/require"
"go.uber.org/zap/zaptest"
"storj.io/common/memory"
"storj.io/common/testcontext"
"storj.io/common/testrand"
"storj.io/common/uuid"
"storj.io/storj/private/testplanet"
"storj.io/storj/satellite/accounting/nodetally"
"storj.io/storj/satellite/audit"
"storj.io/storj/satellite/gc/bloomfilter"
"storj.io/storj/satellite/gracefulexit"
"storj.io/storj/satellite/metabase"
"storj.io/storj/satellite/metabase/metabasetest"
"storj.io/storj/satellite/metabase/rangedloop"
"storj.io/storj/satellite/metabase/rangedloop/rangedlooptest"
"storj.io/storj/satellite/metabase/segmentloop"
"storj.io/storj/satellite/metrics"
"storj.io/storj/satellite/repair/checker"
)
func TestLoopCount(t *testing.T) {
for _, parallelism := range []int{1, 2, 3} {
for _, nSegments := range []int{0, 1, 2, 11} {
for _, nObservers := range []int{0, 1, 2} {
t.Run(
fmt.Sprintf("par%d_seg%d_obs%d", parallelism, nSegments, nObservers),
func(t *testing.T) {
runCountTest(t, parallelism, nSegments, nObservers)
},
)
}
}
}
}
func runCountTest(t *testing.T, parallelism int, nSegments int, nObservers int) {
batchSize := 2
ctx := testcontext.New(t)
observers := []rangedloop.Observer{}
for i := 0; i < nObservers; i++ {
observers = append(observers, &rangedlooptest.CountObserver{})
}
loopService := rangedloop.NewService(
zaptest.NewLogger(t),
rangedloop.Config{
BatchSize: batchSize,
Parallelism: parallelism,
},
&rangedlooptest.RangeSplitter{
Segments: make([]segmentloop.Segment, nSegments),
},
observers,
)
observerDurations, err := loopService.RunOnce(ctx)
require.NoError(t, err)
require.Len(t, observerDurations, nObservers)
for _, observer := range observers {
countObserver := observer.(*rangedlooptest.CountObserver)
require.Equal(t, nSegments, countObserver.NumSegments)
}
}
func TestLoopDuration(t *testing.T) {
t.Skip("Flaky test because it validates concurrency by measuring time")
nSegments := 8
nObservers := 2
parallelism := 4
batchSize := 2
sleepIncrement := time.Millisecond * 10
ctx := testcontext.New(t)
observers := []rangedloop.Observer{}
for i := 0; i < nObservers; i++ {
observers = append(observers, &rangedlooptest.SleepObserver{
Duration: sleepIncrement,
})
}
segments := []segmentloop.Segment{}
for i := 0; i < nSegments; i++ {
streamId, err := uuid.FromBytes([]byte{byte(i), 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0})
require.NoError(t, err)
segments = append(segments, segmentloop.Segment{
StreamID: streamId,
})
}
loopService := rangedloop.NewService(
zaptest.NewLogger(t),
rangedloop.Config{
BatchSize: batchSize,
Parallelism: parallelism,
},
&rangedlooptest.RangeSplitter{
Segments: segments,
},
observers,
)
start := time.Now()
observerDurations, err := loopService.RunOnce(ctx)
require.NoError(t, err)
duration := time.Since(start)
expectedDuration := time.Duration(int64(nSegments) * int64(sleepIncrement) * int64(nObservers) / int64(parallelism))
require.Equal(t, expectedDuration, duration.Truncate(sleepIncrement))
require.Len(t, observerDurations, nObservers)
for _, observerDuration := range observerDurations {
expectedSleep := time.Duration(int64(nSegments) * int64(sleepIncrement))
require.Equal(t, expectedSleep, observerDuration.Duration.Round(sleepIncrement))
}
}
func TestLoopCancellation(t *testing.T) {
parallelism := 2
batchSize := 1
ctx, cancel := context.WithCancel(testcontext.NewWithTimeout(t, time.Second*10))
observers := []rangedloop.Observer{
&rangedlooptest.CountObserver{},
&rangedlooptest.CallbackObserver{
OnProcess: func(ctx context.Context, segments []segmentloop.Segment) error {
// cancel from inside the loop, when it is certain that the loop has started
cancel()
return nil
},
},
}
loopService := rangedloop.NewService(
zaptest.NewLogger(t),
rangedloop.Config{
BatchSize: batchSize,
Parallelism: parallelism,
},
&rangedlooptest.InfiniteSegmentProvider{},
observers,
)
_, err := loopService.RunOnce(ctx)
require.ErrorIs(t, err, context.Canceled)
}
func TestLoopContinuesAfterObserverError(t *testing.T) {
parallelism := 2
batchSize := 1
segments := make([]segmentloop.Segment, 2)
numOnStartCalls := 0
numOnForkCalls := 0
numOnProcessCalls := int32(0)
numOnJoinCalls := 0
numOnFinishCalls := 0
incNumOnProcessCalls := func() {
atomic.AddInt32(&numOnProcessCalls, 1)
}
// first and last observer emit no error
// other observers emit an error at different stages
observers := []rangedloop.Observer{
&rangedlooptest.CallbackObserver{
OnStart: func(ctx context.Context, t time.Time) error {
numOnStartCalls++
return nil
},
OnFork: func(ctx context.Context) (rangedloop.Partial, error) {
numOnForkCalls++
return nil, nil
},
OnProcess: func(ctx context.Context, segments []segmentloop.Segment) error {
incNumOnProcessCalls()
return nil
},
OnJoin: func(ctx context.Context, partial rangedloop.Partial) error {
numOnJoinCalls++
return nil
},
OnFinish: func(ctx context.Context) error {
numOnFinishCalls++
return nil
},
},
&rangedlooptest.CallbackObserver{
OnStart: func(ctx context.Context, t time.Time) error {
numOnStartCalls++
return errors.New("Test OnStart error")
},
OnFork: func(ctx context.Context) (rangedloop.Partial, error) {
require.Fail(t, "OnFork should not be called")
return nil, nil
},
OnProcess: func(ctx context.Context, segments []segmentloop.Segment) error {
require.Fail(t, "OnProcess should not be called")
return nil
},
OnJoin: func(ctx context.Context, partial rangedloop.Partial) error {
require.Fail(t, "OnJoin should not be called")
return nil
},
OnFinish: func(ctx context.Context) error {
require.Fail(t, "OnFinish should not be called")
return nil
},
},
&rangedlooptest.CallbackObserver{
OnStart: func(ctx context.Context, t time.Time) error {
numOnStartCalls++
return nil
},
OnFork: func(ctx context.Context) (rangedloop.Partial, error) {
numOnForkCalls++
return nil, errors.New("Test OnFork error")
},
OnProcess: func(ctx context.Context, segments []segmentloop.Segment) error {
require.Fail(t, "OnProcess should not be called")
return nil
},
OnJoin: func(ctx context.Context, partial rangedloop.Partial) error {
require.Fail(t, "OnJoin should not be called")
return nil
},
OnFinish: func(ctx context.Context) error {
require.Fail(t, "OnFinish should not be called")
return nil
},
},
&rangedlooptest.CallbackObserver{
OnStart: func(ctx context.Context, t time.Time) error {
numOnStartCalls++
return nil
},
OnFork: func(ctx context.Context) (rangedloop.Partial, error) {
numOnForkCalls++
return nil, nil
},
OnProcess: func(ctx context.Context, segments []segmentloop.Segment) error {
incNumOnProcessCalls()
return errors.New("Test OnProcess error")
},
OnJoin: func(ctx context.Context, partial rangedloop.Partial) error {
require.Fail(t, "OnJoin should not be called")
return nil
},
OnFinish: func(ctx context.Context) error {
require.Fail(t, "OnFinish should not be called")
return nil
},
},
&rangedlooptest.CallbackObserver{
OnStart: func(ctx context.Context, t time.Time) error {
numOnStartCalls++
return nil
},
OnFork: func(ctx context.Context) (rangedloop.Partial, error) {
numOnForkCalls++
return nil, nil
},
OnProcess: func(ctx context.Context, segments []segmentloop.Segment) error {
incNumOnProcessCalls()
return nil
},
OnJoin: func(ctx context.Context, partial rangedloop.Partial) error {
numOnJoinCalls++
return errors.New("Test OnJoin error")
},
OnFinish: func(ctx context.Context) error {
require.Fail(t, "OnFinish should not be called")
return nil
},
},
&rangedlooptest.CallbackObserver{
OnStart: func(ctx context.Context, t time.Time) error {
numOnStartCalls++
return nil
},
OnFork: func(ctx context.Context) (rangedloop.Partial, error) {
numOnForkCalls++
return nil, nil
},
OnProcess: func(ctx context.Context, segments []segmentloop.Segment) error {
incNumOnProcessCalls()
return nil
},
OnJoin: func(ctx context.Context, partial rangedloop.Partial) error {
numOnJoinCalls++
return nil
},
OnFinish: func(ctx context.Context) error {
numOnFinishCalls++
return errors.New("Test OnFinish error")
},
},
&rangedlooptest.CallbackObserver{
OnStart: func(ctx context.Context, t time.Time) error {
numOnStartCalls++
return nil
},
OnFork: func(ctx context.Context) (rangedloop.Partial, error) {
numOnForkCalls++
return nil, nil
},
OnProcess: func(ctx context.Context, segments []segmentloop.Segment) error {
incNumOnProcessCalls()
return nil
},
OnJoin: func(ctx context.Context, partial rangedloop.Partial) error {
numOnJoinCalls++
return nil
},
OnFinish: func(ctx context.Context) error {
numOnFinishCalls++
return nil
},
},
}
loopService := rangedloop.NewService(
zaptest.NewLogger(t),
rangedloop.Config{
BatchSize: batchSize,
Parallelism: parallelism,
},
&rangedlooptest.RangeSplitter{
Segments: segments,
},
observers,
)
observerDurations, err := loopService.RunOnce(testcontext.New(t))
require.NoError(t, err)
require.Len(t, observerDurations, len(observers))
require.EqualValues(t, 7, numOnStartCalls)
require.EqualValues(t, 6*parallelism, numOnForkCalls)
require.EqualValues(t, 5*parallelism-1, numOnProcessCalls)
require.EqualValues(t, 4*parallelism-1, numOnJoinCalls)
require.EqualValues(t, 3, numOnFinishCalls)
// success observer should have the duration reported
require.Greater(t, observerDurations[0].Duration, time.Duration(0))
require.Greater(t, observerDurations[6].Duration, time.Duration(0))
// error observers should have sentinel duration reported
require.Equal(t, observerDurations[1].Duration, -1*time.Second)
require.Equal(t, observerDurations[2].Duration, -1*time.Second)
require.Equal(t, observerDurations[3].Duration, -1*time.Second)
require.Equal(t, observerDurations[4].Duration, -1*time.Second)
require.Equal(t, observerDurations[5].Duration, -1*time.Second)
}
func TestAllInOne(t *testing.T) {
testplanet.Run(t, testplanet.Config{
SatelliteCount: 1, StorageNodeCount: 4, UplinkCount: 1,
}, func(t *testing.T, ctx *testcontext.Context, planet *testplanet.Planet) {
log := zaptest.NewLogger(t)
satellite := planet.Satellites[0]
for i := 0; i < 100; i++ {
err := planet.Uplinks[0].Upload(ctx, satellite, "testbucket", "object"+strconv.Itoa(i), testrand.Bytes(5*memory.KiB))
require.NoError(t, err)
}
require.NoError(t, planet.Uplinks[0].CreateBucket(ctx, satellite, "bf-bucket"))
metabaseProvider := rangedloop.NewMetabaseRangeSplitter(satellite.Metabase.DB, 0, 10)
config := rangedloop.Config{
Parallelism: 8,
BatchSize: 3,
}
bfConfig := satellite.Config.GarbageCollectionBF
bfConfig.Bucket = "bf-bucket"
accessGrant, err := planet.Uplinks[0].Access[satellite.ID()].Serialize()
require.NoError(t, err)
bfConfig.AccessGrant = accessGrant
service := rangedloop.NewService(log, config, metabaseProvider, []rangedloop.Observer{
rangedloop.NewLiveCountObserver(satellite.Metabase.DB, config.SuspiciousProcessedRatio, config.AsOfSystemInterval),
metrics.NewObserver(),
nodetally.NewRangedLoopObserver(log.Named("accounting:nodetally"),
satellite.DB.StoragenodeAccounting(),
satellite.Metabase.DB,
),
audit.NewObserver(log.Named("audit"),
satellite.DB.VerifyQueue(),
satellite.Config.Audit,
),
gracefulexit.NewObserver(log.Named("gracefulexit:observer"),
satellite.DB.GracefulExit(),
satellite.DB.OverlayCache(),
satellite.Config.GracefulExit,
),
bloomfilter.NewObserver(log.Named("gc-bf"),
bfConfig,
satellite.DB.OverlayCache(),
),
checker.NewRangedLoopObserver(
log.Named("repair:checker"),
satellite.DB.RepairQueue(),
satellite.Overlay.Service,
satellite.Config.Checker,
),
})
for i := 0; i < 5; i++ {
_, err = service.RunOnce(ctx)
require.NoError(t, err, "iteration %d", i+1)
}
})
}
func TestLoopBoundaries(t *testing.T) {
metabasetest.Run(t, func(ctx *testcontext.Context, t *testing.T, db *metabase.DB) {
type Segment struct {
StreamID uuid.UUID
Position metabase.SegmentPosition
}
var expectedSegments []Segment
parallelism := 4
ranges, err := rangedloop.CreateUUIDRanges(uint32(parallelism))
require.NoError(t, err)
for _, r := range ranges {
if r.Start != nil {
obj := metabasetest.RandObjectStream()
obj.StreamID = *r.Start
metabasetest.CreateObject(ctx, t, db, obj, 1)
expectedSegments = append(expectedSegments, Segment{
StreamID: obj.StreamID,
})
// additional object/segment close to boundary
obj = metabasetest.RandObjectStream()
obj.StreamID = *r.Start
obj.StreamID[len(obj.StreamID)-1]++
metabasetest.CreateObject(ctx, t, db, obj, 1)
expectedSegments = append(expectedSegments, Segment{
StreamID: obj.StreamID,
})
}
}
for _, batchSize := range []int{0, 1, 2, 3, 10} {
var visitedSegments []Segment
var mu sync.Mutex
provider := rangedloop.NewMetabaseRangeSplitter(db, 0, batchSize)
config := rangedloop.Config{
Parallelism: parallelism,
BatchSize: batchSize,
}
callbackObserver := rangedlooptest.CallbackObserver{
OnProcess: func(ctx context.Context, segments []segmentloop.Segment) error {
// OnProcess is called many times by different goroutines
mu.Lock()
defer mu.Unlock()
for _, segment := range segments {
visitedSegments = append(visitedSegments, Segment{
StreamID: segment.StreamID,
Position: segment.Position,
})
}
return nil
},
}
service := rangedloop.NewService(zaptest.NewLogger(t), config, provider, []rangedloop.Observer{&callbackObserver})
_, err = service.RunOnce(ctx)
require.NoError(t, err)
sort.Slice(visitedSegments, func(i, j int) bool {
return visitedSegments[i].StreamID.Less(visitedSegments[j].StreamID)
})
require.Equal(t, expectedSegments, visitedSegments, "batch size %d", batchSize)
}
})
}