2019-06-21 17:27:19 +01:00
|
|
|
// Copyright (C) 2019 Storj Labs, Inc.
|
|
|
|
// See LICENSE for copying information.
|
|
|
|
|
|
|
|
package audit_test
|
|
|
|
|
|
|
|
import (
|
2019-06-24 23:04:06 +01:00
|
|
|
"math/rand"
|
2019-06-21 17:27:19 +01:00
|
|
|
"testing"
|
|
|
|
"time"
|
|
|
|
|
2019-06-24 23:04:06 +01:00
|
|
|
"github.com/stretchr/testify/assert"
|
2019-06-21 17:27:19 +01:00
|
|
|
"github.com/stretchr/testify/require"
|
|
|
|
"go.uber.org/zap"
|
2019-06-24 23:04:06 +01:00
|
|
|
|
|
|
|
"storj.io/storj/internal/memory"
|
2019-06-21 17:27:19 +01:00
|
|
|
"storj.io/storj/internal/testcontext"
|
|
|
|
"storj.io/storj/internal/testplanet"
|
|
|
|
"storj.io/storj/pkg/audit"
|
|
|
|
"storj.io/storj/pkg/overlay"
|
2019-06-24 23:04:06 +01:00
|
|
|
"storj.io/storj/pkg/storage/streams"
|
2019-06-21 17:27:19 +01:00
|
|
|
"storj.io/storj/pkg/storj"
|
|
|
|
"storj.io/storj/satellite"
|
|
|
|
)
|
|
|
|
|
|
|
|
// TestDisqualificationTooManyFailedAudits does the following:
|
|
|
|
// * Create a failed audit report for a storagenode
|
|
|
|
// * Record the audit report several times and check that the node isn't
|
|
|
|
// disqualified until the audit reputation reaches the cut-off value.
|
|
|
|
func TestDisqualificationTooManyFailedAudits(t *testing.T) {
|
|
|
|
var (
|
|
|
|
auditDQCutOff = 0.4
|
|
|
|
alpha0 float64 = 1
|
|
|
|
beta0 float64
|
|
|
|
)
|
|
|
|
|
|
|
|
testplanet.Run(t, testplanet.Config{
|
|
|
|
SatelliteCount: 1, StorageNodeCount: 1, Reconfigure: testplanet.Reconfigure{
|
|
|
|
Satellite: func(log *zap.Logger, index int, config *satellite.Config) {
|
|
|
|
config.Overlay.Node.AuditReputationAlpha0 = alpha0
|
|
|
|
config.Overlay.Node.AuditReputationBeta0 = beta0
|
|
|
|
config.Overlay.Node.AuditReputationLambda = 1
|
|
|
|
config.Overlay.Node.AuditReputationWeight = 1
|
|
|
|
config.Overlay.Node.AuditReputationDQ = auditDQCutOff
|
|
|
|
},
|
|
|
|
},
|
|
|
|
}, func(t *testing.T, ctx *testcontext.Context, planet *testplanet.Planet) {
|
|
|
|
err := planet.Satellites[0].Audit.Service.Close()
|
|
|
|
require.NoError(t, err)
|
|
|
|
|
|
|
|
var (
|
|
|
|
sat = planet.Satellites[0]
|
|
|
|
nodeID = planet.StorageNodes[0].ID()
|
|
|
|
report = &audit.Report{
|
|
|
|
Fails: storj.NodeIDList{nodeID},
|
|
|
|
}
|
|
|
|
)
|
|
|
|
|
|
|
|
dossier, err := sat.Overlay.Service.Get(ctx, nodeID)
|
|
|
|
require.NoError(t, err)
|
|
|
|
|
|
|
|
require.Equal(t, alpha0, dossier.Reputation.AuditReputationAlpha)
|
|
|
|
require.Equal(t, beta0, dossier.Reputation.AuditReputationBeta)
|
|
|
|
|
|
|
|
prevReputation := calcReputation(dossier)
|
|
|
|
|
|
|
|
// Report the audit failure until the node gets disqualified due to many
|
|
|
|
// failed audits
|
|
|
|
iterations := 1
|
|
|
|
for ; ; iterations++ {
|
|
|
|
_, err := sat.Audit.Service.Reporter.RecordAudits(ctx, report)
|
|
|
|
require.NoError(t, err)
|
|
|
|
|
|
|
|
dossier, err := sat.Overlay.Service.Get(ctx, nodeID)
|
|
|
|
require.NoError(t, err)
|
|
|
|
|
|
|
|
reputation := calcReputation(dossier)
|
|
|
|
require.Truef(t, prevReputation >= reputation,
|
|
|
|
"(%d) expected reputation to remain or decrease (previous >= current): %f >= %f",
|
|
|
|
iterations, prevReputation, reputation,
|
|
|
|
)
|
|
|
|
|
|
|
|
if reputation <= auditDQCutOff || reputation == prevReputation {
|
|
|
|
require.NotNilf(t, dossier.Disqualified,
|
|
|
|
"Disqualified (%d) - cut-off: %f, prev. reputation: %f, current reputation: %f",
|
|
|
|
iterations, auditDQCutOff, prevReputation, reputation,
|
|
|
|
)
|
|
|
|
|
|
|
|
require.True(t, time.Now().Sub(*dossier.Disqualified) >= 0,
|
|
|
|
"Disqualified should be in the past",
|
|
|
|
)
|
|
|
|
|
|
|
|
break
|
|
|
|
}
|
|
|
|
|
|
|
|
require.Nil(t, dossier.Disqualified, "Disqualified")
|
|
|
|
prevReputation = reputation
|
|
|
|
}
|
|
|
|
|
|
|
|
require.True(t, iterations > 1, "the number of iterations must be at least 2")
|
|
|
|
})
|
|
|
|
}
|
|
|
|
|
|
|
|
func calcReputation(dossier *overlay.NodeDossier) float64 {
|
|
|
|
var (
|
|
|
|
alpha = dossier.Reputation.AuditReputationAlpha
|
|
|
|
beta = dossier.Reputation.AuditReputationBeta
|
|
|
|
)
|
|
|
|
|
|
|
|
return alpha / (alpha + beta)
|
|
|
|
}
|
2019-06-24 23:04:06 +01:00
|
|
|
|
|
|
|
func TestDisqualifiedNodesGetNoDownload(t *testing.T) {
|
|
|
|
|
|
|
|
// - uploads random data
|
|
|
|
// - mark a node as disqualified
|
|
|
|
// - check we don't get it when we require order limit
|
|
|
|
|
|
|
|
testplanet.Run(t, testplanet.Config{
|
|
|
|
SatelliteCount: 1, StorageNodeCount: 4, UplinkCount: 1,
|
|
|
|
}, func(t *testing.T, ctx *testcontext.Context, planet *testplanet.Planet) {
|
|
|
|
satellite := planet.Satellites[0]
|
|
|
|
upl := planet.Uplinks[0]
|
|
|
|
|
|
|
|
err := satellite.Audit.Service.Close()
|
|
|
|
require.NoError(t, err)
|
|
|
|
|
|
|
|
testData := make([]byte, 8*memory.KiB)
|
|
|
|
_, err = rand.Read(testData)
|
|
|
|
require.NoError(t, err)
|
|
|
|
|
|
|
|
err = upl.Upload(ctx, planet.Satellites[0], "testbucket", "test/path", testData)
|
|
|
|
require.NoError(t, err)
|
|
|
|
|
|
|
|
projects, err := satellite.DB.Console().Projects().GetAll(ctx)
|
|
|
|
require.NoError(t, err)
|
|
|
|
require.Len(t, projects, 1)
|
|
|
|
|
|
|
|
bucketID := []byte(storj.JoinPaths(projects[0].ID.String(), "testbucket"))
|
|
|
|
|
|
|
|
encScheme := upl.GetConfig(satellite).GetEncryptionScheme()
|
|
|
|
cipher := encScheme.Cipher
|
|
|
|
encryptedAfterBucket, err := streams.EncryptAfterBucket(ctx, "testbucket/test/path", cipher, &storj.Key{})
|
|
|
|
require.NoError(t, err)
|
|
|
|
|
|
|
|
lastSegPath := storj.JoinPaths(projects[0].ID.String(), "l", encryptedAfterBucket)
|
|
|
|
pointer, err := satellite.Metainfo.Service.Get(ctx, lastSegPath)
|
|
|
|
require.NoError(t, err)
|
|
|
|
|
|
|
|
disqualifiedNode := pointer.GetRemote().GetRemotePieces()[0].NodeId
|
|
|
|
disqualifyNode(t, ctx, satellite, disqualifiedNode)
|
|
|
|
|
|
|
|
limits, err := satellite.Orders.Service.CreateGetOrderLimits(ctx, upl.Identity.PeerIdentity(), bucketID, pointer)
|
|
|
|
require.NoError(t, err)
|
|
|
|
assert.Len(t, limits, len(pointer.GetRemote().GetRemotePieces())-1)
|
|
|
|
|
|
|
|
for _, orderLimit := range limits {
|
|
|
|
assert.False(t, isDisqualified(t, ctx, satellite, orderLimit.Limit.StorageNodeId))
|
|
|
|
assert.NotEqual(t, orderLimit.Limit.StorageNodeId, disqualifiedNode)
|
|
|
|
}
|
|
|
|
})
|
|
|
|
}
|
|
|
|
|
|
|
|
func TestDisqualifiedNodesGetNoUpload(t *testing.T) {
|
|
|
|
|
|
|
|
// - mark a node as disqualified
|
|
|
|
// - check that we have an error if we try to create a segment using all storage nodes
|
|
|
|
|
|
|
|
testplanet.Run(t, testplanet.Config{
|
|
|
|
SatelliteCount: 1, StorageNodeCount: 4, UplinkCount: 1,
|
|
|
|
}, func(t *testing.T, ctx *testcontext.Context, planet *testplanet.Planet) {
|
|
|
|
satellite := planet.Satellites[0]
|
|
|
|
disqualifiedNode := planet.StorageNodes[0]
|
|
|
|
|
|
|
|
err := satellite.Audit.Service.Close()
|
|
|
|
require.NoError(t, err)
|
|
|
|
|
|
|
|
disqualifyNode(t, ctx, satellite, disqualifiedNode.ID())
|
|
|
|
|
|
|
|
request := overlay.FindStorageNodesRequest{
|
|
|
|
MinimumRequiredNodes: 4,
|
|
|
|
RequestedCount: 0,
|
|
|
|
FreeBandwidth: 0,
|
|
|
|
FreeDisk: 0,
|
|
|
|
ExcludedNodes: nil,
|
|
|
|
MinimumVersion: "", // semver or empty
|
|
|
|
}
|
|
|
|
nodes, err := satellite.Overlay.Service.FindStorageNodes(ctx, request)
|
|
|
|
assert.True(t, overlay.ErrNotEnoughNodes.Has(err))
|
|
|
|
|
|
|
|
assert.Len(t, nodes, 3)
|
|
|
|
for _, node := range nodes {
|
|
|
|
assert.False(t, isDisqualified(t, ctx, satellite, node.Id))
|
|
|
|
assert.NotEqual(t, node.Id, disqualifiedNode)
|
|
|
|
}
|
|
|
|
|
|
|
|
})
|
|
|
|
}
|
|
|
|
|
|
|
|
func isDisqualified(t *testing.T, ctx *testcontext.Context, satellite *satellite.Peer, nodeID storj.NodeID) bool {
|
|
|
|
node, err := satellite.Overlay.Service.Get(ctx, nodeID)
|
|
|
|
require.NoError(t, err)
|
|
|
|
|
|
|
|
return node.Disqualified != nil
|
|
|
|
}
|
|
|
|
func disqualifyNode(t *testing.T, ctx *testcontext.Context, satellite *satellite.Peer, nodeID storj.NodeID) {
|
|
|
|
_, err := satellite.DB.OverlayCache().UpdateStats(ctx, &overlay.UpdateRequest{
|
|
|
|
NodeID: nodeID,
|
|
|
|
IsUp: true,
|
|
|
|
AuditSuccess: false,
|
|
|
|
AuditLambda: 0,
|
|
|
|
AuditWeight: 1,
|
|
|
|
AuditDQ: 0.5,
|
|
|
|
UptimeLambda: 1,
|
|
|
|
UptimeWeight: 1,
|
|
|
|
UptimeDQ: 0.5,
|
|
|
|
})
|
|
|
|
require.NoError(t, err)
|
|
|
|
assert.True(t, isDisqualified(t, ctx, satellite, nodeID))
|
|
|
|
}
|