VictoriaMetrics/lib/storage/table_search_test.go
Aliaksandr Valialkin 52006149b2
lib/storage: replace OpenStorage() with MustOpenStorage()
Callers of OpenStorage() log the returned error and exit.
The error logging and exit can be performed inside MustOpenStorage()
alongside with printing the stack trace for better debuggability.
This simplifies the code at caller side.
2023-04-14 23:02:40 -07:00

283 lines
7.7 KiB
Go

package storage
import (
"fmt"
"math/rand"
"os"
"sort"
"testing"
"time"
)
func TestTableSearch(t *testing.T) {
rng := rand.New(rand.NewSource(1))
var trData TimeRange
trData.fromPartitionTime(time.Now())
trData.MinTimestamp -= 5 * 365 * 24 * 3600 * 1000
t.Run("SinglePartition", func(t *testing.T) {
trSearch := TimeRange{
MinTimestamp: trData.MinTimestamp + 4e3,
MaxTimestamp: trData.MaxTimestamp - 4e3,
}
testTableSearchEx(t, rng, trData, trSearch, 1, 10, 1000, 10)
})
t.Run("SinglePartPerPartition", func(t *testing.T) {
trSearch := TimeRange{
MinTimestamp: trData.MinTimestamp + 4e3,
MaxTimestamp: trData.MaxTimestamp - 4e3,
}
testTableSearchEx(t, rng, trData, trSearch, 12, 1, 1000, 10)
})
t.Run("SingleRowPerPartition", func(t *testing.T) {
trSearch := TimeRange{
MinTimestamp: trData.MinTimestamp + 4e3,
MaxTimestamp: trData.MaxTimestamp - 4e3,
}
testTableSearchEx(t, rng, trData, trSearch, 12, 20, 1, 10)
})
t.Run("SingleTSID", func(t *testing.T) {
trSearch := TimeRange{
MinTimestamp: trData.MinTimestamp + 4e3,
MaxTimestamp: trData.MaxTimestamp - 4e3,
}
testTableSearchEx(t, rng, trData, trSearch, 12, 5, 1000, 1)
})
t.Run("ManyPartitions", func(t *testing.T) {
trSearch := TimeRange{
MinTimestamp: trData.MinTimestamp + 4e3,
MaxTimestamp: trData.MaxTimestamp - 4e3,
}
testTableSearchEx(t, rng, trData, trSearch, 20, 10, 30, 20)
})
t.Run("ManyTSIDs", func(t *testing.T) {
trSearch := TimeRange{
MinTimestamp: trData.MinTimestamp + 4e3,
MaxTimestamp: trData.MaxTimestamp - 4e3,
}
testTableSearchEx(t, rng, trData, trSearch, 2, 5, 5000, 1000)
})
t.Run("ExactTimeRange", func(t *testing.T) {
trSearch := TimeRange{
MinTimestamp: trData.MinTimestamp,
MaxTimestamp: trData.MaxTimestamp,
}
testTableSearchEx(t, rng, trData, trSearch, 2, 5, 1000, 10)
})
t.Run("InnerTimeRange", func(t *testing.T) {
trSearch := TimeRange{
MinTimestamp: trData.MinTimestamp + 4e3,
MaxTimestamp: trData.MaxTimestamp - 4e3,
}
testTableSearchEx(t, rng, trData, trSearch, 2, 5, 1000, 10)
})
t.Run("OuterTimeRange", func(t *testing.T) {
trSearch := TimeRange{
MinTimestamp: trData.MinTimestamp - 1e6,
MaxTimestamp: trData.MaxTimestamp + 1e6,
}
testTableSearchEx(t, rng, trData, trSearch, 2, 5, 1000, 10)
})
t.Run("LowTimeRange", func(t *testing.T) {
trSearch := TimeRange{
MinTimestamp: trData.MinTimestamp - 2e6,
MaxTimestamp: trData.MinTimestamp - 1e6,
}
testTableSearchEx(t, rng, trData, trSearch, 2, 5, 1000, 10)
})
t.Run("HighTimeRange", func(t *testing.T) {
trSearch := TimeRange{
MinTimestamp: trData.MaxTimestamp + 1e6,
MaxTimestamp: trData.MaxTimestamp + 2e6,
}
testTableSearchEx(t, rng, trData, trSearch, 2, 5, 1000, 10)
})
t.Run("LowerEndTimeRange", func(t *testing.T) {
trSearch := TimeRange{
MinTimestamp: trData.MinTimestamp - 1e6,
MaxTimestamp: trData.MaxTimestamp - 4e3,
}
testTableSearchEx(t, rng, trData, trSearch, 2, 5, 1000, 10)
})
t.Run("HigherEndTimeRange", func(t *testing.T) {
trSearch := TimeRange{
MinTimestamp: trData.MinTimestamp + 4e3,
MaxTimestamp: trData.MaxTimestamp + 1e6,
}
testTableSearchEx(t, rng, trData, trSearch, 2, 5, 1000, 10)
})
}
func testTableSearchEx(t *testing.T, rng *rand.Rand, trData, trSearch TimeRange, partitionsCount, maxPartsPerPartition, maxRowsPerPart, tsidsCount int) {
t.Helper()
// Generate tsids to search.
var tsids []TSID
var tsid TSID
for i := 0; i < 25; i++ {
tsid.MetricID = uint64(rng.Intn(tsidsCount * 2))
tsids = append(tsids, tsid)
}
sort.Slice(tsids, func(i, j int) bool { return tsids[i].Less(&tsids[j]) })
// Generate the expected blocks.
var r rawRow
r.PrecisionBits = 24
rowsCountExpected := int64(0)
rbsExpected := []rawBlock{}
var ptr TimeRange
ptr.fromPartitionTimestamp(trData.MinTimestamp)
var rowss [][]rawRow
for i := 0; i < partitionsCount; i++ {
partsCount := rng.Intn(maxPartsPerPartition) + 1
for j := 0; j < partsCount; j++ {
var rows []rawRow
timestamp := ptr.MinTimestamp
rowsCount := rng.Intn(maxRowsPerPart) + 1
for k := 0; k < rowsCount; k++ {
r.TSID.MetricID = uint64(rng.Intn(tsidsCount))
r.Timestamp = timestamp
r.Value = float64(int(rng.NormFloat64() * 1e5))
timestamp += int64(rng.Intn(1e4)) + 1
if timestamp > ptr.MaxTimestamp {
break
}
rows = append(rows, r)
rowsCountExpected++
}
rbs := getTestExpectedRawBlocks(rows, tsids, trSearch)
rbsExpected = append(rbsExpected, rbs...)
rowss = append(rowss, rows)
}
// Go to the next partition.
ptr.fromPartitionTimestamp(ptr.MaxTimestamp + 1)
if ptr.MaxTimestamp > trData.MaxTimestamp {
break
}
}
sort.Slice(rbsExpected, func(i, j int) bool {
a, b := rbsExpected[i], rbsExpected[j]
if a.TSID.Less(&b.TSID) {
return true
}
if b.TSID.Less(&a.TSID) {
return false
}
return a.Timestamps[0] < b.Timestamps[0]
})
// Create a table from rowss and test search on it.
strg := newTestStorage()
tb := mustOpenTable("test-table", strg)
defer func() {
if err := os.RemoveAll("test-table"); err != nil {
t.Fatalf("cannot remove table directory: %s", err)
}
}()
for _, rows := range rowss {
tb.MustAddRows(rows)
// Flush rows to parts.
tb.flushPendingRows()
}
testTableSearch(t, tb, tsids, trSearch, rbsExpected, -1)
tb.MustClose()
// Open the created table and test search on it.
tb = mustOpenTable("test-table", strg)
testTableSearch(t, tb, tsids, trSearch, rbsExpected, rowsCountExpected)
tb.MustClose()
}
func testTableSearch(t *testing.T, tb *table, tsids []TSID, tr TimeRange, rbsExpected []rawBlock, rowsCountExpected int64) {
t.Helper()
if err := testTableSearchSerial(tb, tsids, tr, rbsExpected, rowsCountExpected); err != nil {
t.Fatalf("unexpected error in serial table search: %s", err)
}
ch := make(chan error, 5)
for i := 0; i < cap(ch); i++ {
go func() {
ch <- testTableSearchSerial(tb, tsids, tr, rbsExpected, rowsCountExpected)
}()
}
for i := 0; i < cap(ch); i++ {
select {
case err := <-ch:
if err != nil {
t.Fatalf("unexpected error in concurrent table search: %s", err)
}
case <-time.After(time.Second):
t.Fatalf("timeout in concurrent table search")
}
}
}
func testTableSearchSerial(tb *table, tsids []TSID, tr TimeRange, rbsExpected []rawBlock, rowsCountExpected int64) error {
if rowsCountExpected >= 0 {
// Verify rows count only on the table opened from file.
//
// The online created table may not contain all the rows, since
// they may race with raw rows flusher.
var m TableMetrics
tb.UpdateMetrics(&m)
if rowsCount := m.TotalRowsCount(); rowsCount != uint64(rowsCountExpected) {
return fmt.Errorf("unexpected rows count in the table; got %d; want %d", rowsCount, rowsCountExpected)
}
}
bs := []Block{}
var ts tableSearch
ts.Init(tb, tsids, tr)
for ts.NextBlock() {
var b Block
ts.BlockRef.MustReadBlock(&b)
bs = append(bs, b)
}
if err := ts.Error(); err != nil {
return fmt.Errorf("unexpected error: %w", err)
}
ts.MustClose()
rbs := newTestRawBlocks(bs, tr)
if err := testEqualRawBlocks(rbs, rbsExpected); err != nil {
return fmt.Errorf("unequal blocks: %w", err)
}
if rowsCountExpected >= 0 {
var m TableMetrics
tb.UpdateMetrics(&m)
if rowsCount := m.TotalRowsCount(); rowsCount != uint64(rowsCountExpected) {
return fmt.Errorf("unexpected rows count in the table; got %d; want %d", rowsCount, rowsCountExpected)
}
}
// verify that empty tsids returns empty result
ts.Init(tb, []TSID{}, tr)
if ts.NextBlock() {
return fmt.Errorf("unexpected block got for an empty tsids list: %+v", ts.BlockRef)
}
if err := ts.Error(); err != nil {
return fmt.Errorf("unexpected error on empty tsids list: %w", err)
}
ts.MustClose()
return nil
}