minio/cmd/xl-v1-utils_test.go
Harshavardhana d45a8784fc
Fix hash order to generate more even distribution (#5247)
The problem in existing code was the following line

```
start := int(keyCrc%uint32(cardinality)) | 1
```

A given a value of N cardinality the ending result
because of the the bitwise '|' would lead to always
higher affinity to odd sequences.

As can be seen from the test cases that this can
lead to many objects being allocated the same set
of disks or atleast the first disk is an odd disk
always.  This introduces a performance problem
for majority of the objects under concurrent load.

Remove `| 1` to provide a more cleaner distribution
and the new code will be.
```
start := int(keyCrc % uint32(cardinality))
```

Thanks to Krishna Srinivas for pointing out the bitwise
situation here.
2017-11-30 12:57:03 -08:00

452 lines
17 KiB
Go

/*
* Minio Cloud Storage, (C) 2015, 2016, 2017 Minio, Inc.
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package cmd
import (
"bytes"
"encoding/hex"
"encoding/json"
"reflect"
"strconv"
"testing"
humanize "github.com/dustin/go-humanize"
"github.com/minio/minio/pkg/errors"
)
// Tests caclculating disk count.
func TestDiskCount(t *testing.T) {
testCases := []struct {
disks []StorageAPI
diskCount int
}{
// Test case - 1
{
disks: []StorageAPI{&posix{}, &posix{}, &posix{}, &posix{}},
diskCount: 4,
},
// Test case - 2
{
disks: []StorageAPI{nil, &posix{}, &posix{}, &posix{}},
diskCount: 3,
},
}
for i, testCase := range testCases {
cdiskCount := diskCount(testCase.disks)
if cdiskCount != testCase.diskCount {
t.Errorf("Test %d: Expected %d, got %d", i+1, testCase.diskCount, cdiskCount)
}
}
}
// Test for reduceErrs, reduceErr reduces collection
// of errors into a single maximal error with in the list.
func TestReduceErrs(t *testing.T) {
// List all of all test cases to validate various cases of reduce errors.
testCases := []struct {
errs []error
ignoredErrs []error
err error
}{
// Validate if have reduced properly.
{[]error{
errDiskNotFound,
errDiskNotFound,
errDiskFull,
}, []error{}, errXLReadQuorum},
// Validate if have no consensus.
{[]error{
errDiskFull,
errDiskNotFound,
nil, nil,
}, []error{}, errXLReadQuorum},
// Validate if have consensus and errors ignored.
{[]error{
errVolumeNotFound,
errVolumeNotFound,
errVolumeNotFound,
errVolumeNotFound,
errVolumeNotFound,
errDiskNotFound,
errDiskNotFound,
}, []error{errDiskNotFound}, errVolumeNotFound},
{[]error{}, []error{}, errXLReadQuorum},
{[]error{errFileNotFound, errFileNotFound, errFileNotFound,
errFileNotFound, errFileNotFound, nil, nil, nil, nil, nil},
nil, nil},
}
// Validates list of all the testcases for returning valid errors.
for i, testCase := range testCases {
gotErr := reduceReadQuorumErrs(testCase.errs, testCase.ignoredErrs, 5)
if errors.Cause(gotErr) != testCase.err {
t.Errorf("Test %d : expected %s, got %s", i+1, testCase.err, gotErr)
}
gotNewErr := reduceWriteQuorumErrs(testCase.errs, testCase.ignoredErrs, 6)
if errors.Cause(gotNewErr) != errXLWriteQuorum {
t.Errorf("Test %d : expected %s, got %s", i+1, errXLWriteQuorum, gotErr)
}
}
}
// TestHashOrder - test order of ints in array
func TestHashOrder(t *testing.T) {
testCases := []struct {
objectName string
hashedOrder []int
}{
// cases which should pass the test.
// passing in valid object name.
{"object", []int{14, 15, 16, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13}},
{"The Shining Script <v1>.pdf", []int{16, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15}},
{"Cost Benefit Analysis (2009-2010).pptx", []int{15, 16, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14}},
{"117Gn8rfHL2ACARPAhaFd0AGzic9pUbIA/5OCn5A", []int{3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 1, 2}},
{"SHØRT", []int{11, 12, 13, 14, 15, 16, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10}},
{"There are far too many object names, and far too few bucket names!", []int{15, 16, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14}},
{"a/b/c/", []int{3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 1, 2}},
{"/a/b/c", []int{6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 1, 2, 3, 4, 5}},
{string([]byte{0xff, 0xfe, 0xfd}), []int{15, 16, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14}},
}
// Tests hashing order to be consistent.
for i, testCase := range testCases {
hashedOrder := hashOrder(testCase.objectName, 16)
if !reflect.DeepEqual(testCase.hashedOrder, hashedOrder) {
t.Errorf("Test case %d: Expected \"%#v\" but failed \"%#v\"", i+1, testCase.hashedOrder, hashedOrder)
}
}
// Tests hashing order to fail for when order is '-1'.
if hashedOrder := hashOrder("This will fail", -1); hashedOrder != nil {
t.Errorf("Test: Expect \"nil\" but failed \"%#v\"", hashedOrder)
}
}
// newTestXLMetaV1 - initializes new xlMetaV1, adds version, allocates a fresh erasure info and metadata.
func newTestXLMetaV1() xlMetaV1 {
xlMeta := xlMetaV1{}
xlMeta.Version = xlMetaVersion
xlMeta.Format = xlMetaFormat
xlMeta.Minio.Release = "test"
xlMeta.Erasure = ErasureInfo{
Algorithm: "klauspost/reedsolomon/vandermonde",
DataBlocks: 5,
ParityBlocks: 5,
BlockSize: 10485760,
Index: 10,
Distribution: []int{9, 10, 1, 2, 3, 4, 5, 6, 7, 8},
}
xlMeta.Stat = statInfo{
Size: int64(20),
ModTime: UTCNow(),
}
// Set meta data.
xlMeta.Meta = make(map[string]string)
xlMeta.Meta["testKey1"] = "val1"
xlMeta.Meta["testKey2"] = "val2"
return xlMeta
}
func (m *xlMetaV1) AddTestObjectCheckSum(checkSumNum int, name string, algorithm BitrotAlgorithm, hash string) {
checksum, err := hex.DecodeString(hash)
if err != nil {
panic(err)
}
m.Erasure.Checksums[checkSumNum] = ChecksumInfo{name, algorithm, checksum}
}
// AddTestObjectPart - add a new object part in order.
func (m *xlMetaV1) AddTestObjectPart(partNumber int, partName string, partETag string, partSize int64) {
partInfo := objectPartInfo{
Number: partNumber,
Name: partName,
ETag: partETag,
Size: partSize,
}
// Proceed to include new part info.
m.Parts[partNumber] = partInfo
}
// Constructs xlMetaV1{} for given number of parts and converts it into bytes.
func getXLMetaBytes(totalParts int) []byte {
xlSampleMeta := getSampleXLMeta(totalParts)
xlMetaBytes, err := json.Marshal(xlSampleMeta)
if err != nil {
panic(err)
}
return xlMetaBytes
}
// Returns sample xlMetaV1{} for number of parts.
func getSampleXLMeta(totalParts int) xlMetaV1 {
xlMeta := newTestXLMetaV1()
// Number of checksum info == total parts.
xlMeta.Erasure.Checksums = make([]ChecksumInfo, totalParts)
// total number of parts.
xlMeta.Parts = make([]objectPartInfo, totalParts)
for i := 0; i < totalParts; i++ {
partName := "part." + strconv.Itoa(i+1)
// hard coding hash and algo value for the checksum, Since we are benchmarking the parsing of xl.json the magnitude doesn't affect the test,
// The magnitude doesn't make a difference, only the size does.
xlMeta.AddTestObjectCheckSum(i, partName, BLAKE2b512, "a23f5eff248c4372badd9f3b2455a285cd4ca86c3d9a570b091d3fc5cd7ca6d9484bbea3f8c5d8d4f84daae96874419eda578fd736455334afbac2c924b3915a")
xlMeta.AddTestObjectPart(i, partName, "d3fdd79cc3efd5fe5c068d7be397934b", 67108864)
}
return xlMeta
}
// Compare the unmarshaled XLMetaV1 with the one obtained from gjson parsing.
func compareXLMetaV1(t *testing.T, unMarshalXLMeta, gjsonXLMeta xlMetaV1) {
// Start comparing the fields of xlMetaV1 obtained from gjson parsing with one parsed using json unmarshaling.
if unMarshalXLMeta.Version != gjsonXLMeta.Version {
t.Errorf("Expected the Version to be \"%s\", but got \"%s\".", unMarshalXLMeta.Version, gjsonXLMeta.Version)
}
if unMarshalXLMeta.Format != gjsonXLMeta.Format {
t.Errorf("Expected the format to be \"%s\", but got \"%s\".", unMarshalXLMeta.Format, gjsonXLMeta.Format)
}
if unMarshalXLMeta.Stat.Size != gjsonXLMeta.Stat.Size {
t.Errorf("Expected the stat size to be %v, but got %v.", unMarshalXLMeta.Stat.Size, gjsonXLMeta.Stat.Size)
}
if !unMarshalXLMeta.Stat.ModTime.Equal(gjsonXLMeta.Stat.ModTime) {
t.Errorf("Expected the modTime to be \"%v\", but got \"%v\".", unMarshalXLMeta.Stat.ModTime, gjsonXLMeta.Stat.ModTime)
}
if unMarshalXLMeta.Erasure.Algorithm != gjsonXLMeta.Erasure.Algorithm {
t.Errorf("Expected the erasure algorithm to be \"%v\", but got \"%v\".", unMarshalXLMeta.Erasure.Algorithm, gjsonXLMeta.Erasure.Algorithm)
}
if unMarshalXLMeta.Erasure.DataBlocks != gjsonXLMeta.Erasure.DataBlocks {
t.Errorf("Expected the erasure data blocks to be %v, but got %v.", unMarshalXLMeta.Erasure.DataBlocks, gjsonXLMeta.Erasure.DataBlocks)
}
if unMarshalXLMeta.Erasure.ParityBlocks != gjsonXLMeta.Erasure.ParityBlocks {
t.Errorf("Expected the erasure parity blocks to be %v, but got %v.", unMarshalXLMeta.Erasure.ParityBlocks, gjsonXLMeta.Erasure.ParityBlocks)
}
if unMarshalXLMeta.Erasure.BlockSize != gjsonXLMeta.Erasure.BlockSize {
t.Errorf("Expected the erasure block size to be %v, but got %v.", unMarshalXLMeta.Erasure.BlockSize, gjsonXLMeta.Erasure.BlockSize)
}
if unMarshalXLMeta.Erasure.Index != gjsonXLMeta.Erasure.Index {
t.Errorf("Expected the erasure index to be %v, but got %v.", unMarshalXLMeta.Erasure.Index, gjsonXLMeta.Erasure.Index)
}
if len(unMarshalXLMeta.Erasure.Distribution) != len(gjsonXLMeta.Erasure.Distribution) {
t.Errorf("Expected the size of Erasure Distribution to be %d, but got %d.", len(unMarshalXLMeta.Erasure.Distribution), len(gjsonXLMeta.Erasure.Distribution))
} else {
for i := 0; i < len(unMarshalXLMeta.Erasure.Distribution); i++ {
if unMarshalXLMeta.Erasure.Distribution[i] != gjsonXLMeta.Erasure.Distribution[i] {
t.Errorf("Expected the Erasure Distribution to be %d, got %d.", unMarshalXLMeta.Erasure.Distribution[i], gjsonXLMeta.Erasure.Distribution[i])
}
}
}
if len(unMarshalXLMeta.Erasure.Checksums) != len(gjsonXLMeta.Erasure.Checksums) {
t.Errorf("Expected the size of Erasure Checksums to be %d, but got %d.", len(unMarshalXLMeta.Erasure.Checksums), len(gjsonXLMeta.Erasure.Checksums))
} else {
for i := 0; i < len(unMarshalXLMeta.Erasure.Checksums); i++ {
if unMarshalXLMeta.Erasure.Checksums[i].Name != gjsonXLMeta.Erasure.Checksums[i].Name {
t.Errorf("Expected the Erasure Checksum Name to be \"%s\", got \"%s\".", unMarshalXLMeta.Erasure.Checksums[i].Name, gjsonXLMeta.Erasure.Checksums[i].Name)
}
if unMarshalXLMeta.Erasure.Checksums[i].Algorithm != gjsonXLMeta.Erasure.Checksums[i].Algorithm {
t.Errorf("Expected the Erasure Checksum Algorithm to be \"%s\", got \"%s\".", unMarshalXLMeta.Erasure.Checksums[i].Algorithm, gjsonXLMeta.Erasure.Checksums[i].Algorithm)
}
if !bytes.Equal(unMarshalXLMeta.Erasure.Checksums[i].Hash, gjsonXLMeta.Erasure.Checksums[i].Hash) {
t.Errorf("Expected the Erasure Checksum Hash to be \"%s\", got \"%s\".", unMarshalXLMeta.Erasure.Checksums[i].Hash, gjsonXLMeta.Erasure.Checksums[i].Hash)
}
}
}
if unMarshalXLMeta.Minio.Release != gjsonXLMeta.Minio.Release {
t.Errorf("Expected the Release string to be \"%s\", but got \"%s\".", unMarshalXLMeta.Minio.Release, gjsonXLMeta.Minio.Release)
}
if len(unMarshalXLMeta.Parts) != len(gjsonXLMeta.Parts) {
t.Errorf("Expected info of %d parts to be present, but got %d instead.", len(unMarshalXLMeta.Parts), len(gjsonXLMeta.Parts))
} else {
for i := 0; i < len(unMarshalXLMeta.Parts); i++ {
if unMarshalXLMeta.Parts[i].Name != gjsonXLMeta.Parts[i].Name {
t.Errorf("Expected the name of part %d to be \"%s\", got \"%s\".", i+1, unMarshalXLMeta.Parts[i].Name, gjsonXLMeta.Parts[i].Name)
}
if unMarshalXLMeta.Parts[i].ETag != gjsonXLMeta.Parts[i].ETag {
t.Errorf("Expected the ETag of part %d to be \"%s\", got \"%s\".", i+1, unMarshalXLMeta.Parts[i].ETag, gjsonXLMeta.Parts[i].ETag)
}
if unMarshalXLMeta.Parts[i].Number != gjsonXLMeta.Parts[i].Number {
t.Errorf("Expected the number of part %d to be \"%d\", got \"%d\".", i+1, unMarshalXLMeta.Parts[i].Number, gjsonXLMeta.Parts[i].Number)
}
if unMarshalXLMeta.Parts[i].Size != gjsonXLMeta.Parts[i].Size {
t.Errorf("Expected the size of part %d to be %v, got %v.", i+1, unMarshalXLMeta.Parts[i].Size, gjsonXLMeta.Parts[i].Size)
}
}
}
for key, val := range unMarshalXLMeta.Meta {
gjsonVal, exists := gjsonXLMeta.Meta[key]
if !exists {
t.Errorf("No meta data entry for Key \"%s\" exists.", key)
}
if val != gjsonVal {
t.Errorf("Expected the value for Meta data key \"%s\" to be \"%s\", but got \"%s\".", key, val, gjsonVal)
}
}
}
// Tests the correctness of constructing XLMetaV1 using gjson lib.
// The result will be compared with the result obtained from json.unMarshal of the byte data.
func TestGetXLMetaV1GJson1(t *testing.T) {
xlMetaJSON := getXLMetaBytes(1)
var unMarshalXLMeta xlMetaV1
if err := json.Unmarshal(xlMetaJSON, &unMarshalXLMeta); err != nil {
t.Errorf("Unmarshalling failed: %v", err)
}
gjsonXLMeta, err := xlMetaV1UnmarshalJSON(xlMetaJSON)
if err != nil {
t.Errorf("gjson parsing of XLMeta failed: %v", err)
}
compareXLMetaV1(t, unMarshalXLMeta, gjsonXLMeta)
}
// Tests the correctness of constructing XLMetaV1 using gjson lib for XLMetaV1 of size 10 parts.
// The result will be compared with the result obtained from json.unMarshal of the byte data.
func TestGetXLMetaV1GJson10(t *testing.T) {
xlMetaJSON := getXLMetaBytes(10)
var unMarshalXLMeta xlMetaV1
if err := json.Unmarshal(xlMetaJSON, &unMarshalXLMeta); err != nil {
t.Errorf("Unmarshalling failed: %v", err)
}
gjsonXLMeta, err := xlMetaV1UnmarshalJSON(xlMetaJSON)
if err != nil {
t.Errorf("gjson parsing of XLMeta failed: %v", err)
}
compareXLMetaV1(t, unMarshalXLMeta, gjsonXLMeta)
}
// Test the predicted part size from the part index
func TestGetPartSizeFromIdx(t *testing.T) {
// Create test cases
testCases := []struct {
totalSize int64
partSize int64
partIndex int
expectedSize int64
}{
// Total size is zero
{0, 10, 1, 0},
// part size 2MiB, total size 4MiB
{4 * humanize.MiByte, 2 * humanize.MiByte, 1, 2 * humanize.MiByte},
{4 * humanize.MiByte, 2 * humanize.MiByte, 2, 2 * humanize.MiByte},
{4 * humanize.MiByte, 2 * humanize.MiByte, 3, 0},
// part size 2MiB, total size 5MiB
{5 * humanize.MiByte, 2 * humanize.MiByte, 1, 2 * humanize.MiByte},
{5 * humanize.MiByte, 2 * humanize.MiByte, 2, 2 * humanize.MiByte},
{5 * humanize.MiByte, 2 * humanize.MiByte, 3, 1 * humanize.MiByte},
{5 * humanize.MiByte, 2 * humanize.MiByte, 4, 0},
}
for i, testCase := range testCases {
s, err := calculatePartSizeFromIdx(testCase.totalSize, testCase.partSize, testCase.partIndex)
if err != nil {
t.Errorf("Test %d: Expected to pass but failed. %s", i+1, err)
}
if err == nil && s != testCase.expectedSize {
t.Errorf("Test %d: The calculated part size is incorrect: expected = %d, found = %d\n", i+1, testCase.expectedSize, s)
}
}
testCasesFailure := []struct {
totalSize int64
partSize int64
partIndex int
err error
}{
// partSize is 0, returns error.
{10, 0, 1, errPartSizeZero},
// partIndex is 0, returns error.
{10, 1, 0, errPartSizeIndex},
// Total size is -1, returns error.
{-1, 10, 1, errInvalidArgument},
}
for i, testCaseFailure := range testCasesFailure {
_, err := calculatePartSizeFromIdx(testCaseFailure.totalSize, testCaseFailure.partSize, testCaseFailure.partIndex)
if err == nil {
t.Errorf("Test %d: Expected to failed but passed. %s", i+1, err)
}
if err != nil && errors.Cause(err) != testCaseFailure.err {
t.Errorf("Test %d: Expected err %s, but got %s", i+1, testCaseFailure.err, errors.Cause(err))
}
}
}
func TestShuffleDisks(t *testing.T) {
nDisks := 16
disks, err := getRandomDisks(nDisks)
if err != nil {
t.Fatal(err)
}
objLayer, _, err := initObjectLayer(mustGetNewEndpointList(disks...))
if err != nil {
removeRoots(disks)
t.Fatal(err)
}
defer removeRoots(disks)
xl := objLayer.(*xlObjects)
testShuffleDisks(t, xl)
}
// Test shuffleDisks which returns shuffled slice of disks for their actual distribution.
func testShuffleDisks(t *testing.T, xl *xlObjects) {
disks := xl.storageDisks
distribution := []int{16, 14, 12, 10, 8, 6, 4, 2, 1, 3, 5, 7, 9, 11, 13, 15}
shuffledDisks := shuffleDisks(disks, distribution)
// From the "distribution" above you can notice that:
// 1st data block is in the 9th disk (i.e distribution index 8)
// 2nd data block is in the 8th disk (i.e distribution index 7) and so on.
if shuffledDisks[0] != disks[8] ||
shuffledDisks[1] != disks[7] ||
shuffledDisks[2] != disks[9] ||
shuffledDisks[3] != disks[6] ||
shuffledDisks[4] != disks[10] ||
shuffledDisks[5] != disks[5] ||
shuffledDisks[6] != disks[11] ||
shuffledDisks[7] != disks[4] ||
shuffledDisks[8] != disks[12] ||
shuffledDisks[9] != disks[3] ||
shuffledDisks[10] != disks[13] ||
shuffledDisks[11] != disks[2] ||
shuffledDisks[12] != disks[14] ||
shuffledDisks[13] != disks[1] ||
shuffledDisks[14] != disks[15] ||
shuffledDisks[15] != disks[0] {
t.Errorf("shuffleDisks returned incorrect order.")
}
}
// TestEvalDisks tests the behavior of evalDisks
func TestEvalDisks(t *testing.T) {
nDisks := 16
disks, err := getRandomDisks(nDisks)
if err != nil {
t.Fatal(err)
}
objLayer, _, err := initObjectLayer(mustGetNewEndpointList(disks...))
if err != nil {
removeRoots(disks)
t.Fatal(err)
}
defer removeRoots(disks)
xl := objLayer.(*xlObjects)
testShuffleDisks(t, xl)
}