1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
|
package integration
import (
"fmt"
"sync"
"sync/atomic"
"testing"
"time"
"github.com/seaweedfs/seaweedfs/weed/glog"
)
// TestResumeMillionRecords_Fixed - Fixed version with better concurrency handling
func TestResumeMillionRecords_Fixed(t *testing.T) {
const (
totalRecords = 1000000
numPartitions = int32(8)
numProducers = 4
brokerAddr = "localhost:17777"
batchSize = 100 // Process in smaller batches to avoid overwhelming
)
// Create direct broker client
client, err := NewDirectBrokerClient(brokerAddr)
if err != nil {
t.Fatalf("Failed to create direct broker client: %v", err)
}
defer client.Close()
topicName := fmt.Sprintf("resume-million-test-%d", time.Now().Unix())
// Create topic
glog.Infof("Creating topic %s with %d partitions for RESUMED test", topicName, numPartitions)
err = client.ConfigureTopic(topicName, numPartitions)
if err != nil {
t.Fatalf("Failed to configure topic: %v", err)
}
// Performance tracking
var totalProduced int64
var totalErrors int64
startTime := time.Now()
// Progress tracking
ticker := time.NewTicker(5 * time.Second) // More frequent updates
defer ticker.Stop()
go func() {
for range ticker.C {
produced := atomic.LoadInt64(&totalProduced)
errors := atomic.LoadInt64(&totalErrors)
elapsed := time.Since(startTime)
rate := float64(produced) / elapsed.Seconds()
progressPercent := float64(produced) / float64(totalRecords) * 100
glog.Infof("PROGRESS: %d/%d records (%.1f%%), rate: %.0f records/sec, errors: %d",
produced, totalRecords, progressPercent, rate, errors)
if produced >= totalRecords {
return
}
}
}()
// Fixed producer function with better error handling
producer := func(producerID int, recordsPerProducer int) error {
defer glog.Infof("Producer %d FINISHED", producerID)
// Create dedicated clients per producer to avoid contention
producerClient, err := NewDirectBrokerClient(brokerAddr)
if err != nil {
return fmt.Errorf("producer %d failed to create client: %v", producerID, err)
}
defer producerClient.Close()
successCount := 0
for i := 0; i < recordsPerProducer; i++ {
recordID := producerID*recordsPerProducer + i
// Generate test record
testRecord := GenerateMockTestRecord(recordID)
key, value := SerializeMockTestRecord(testRecord)
partition := int32(testRecord.UserID % int64(numPartitions))
// Produce with retry logic
maxRetries := 3
var lastErr error
success := false
for retry := 0; retry < maxRetries; retry++ {
err := producerClient.PublishRecord(topicName, partition, key, value)
if err == nil {
success = true
break
}
lastErr = err
time.Sleep(time.Duration(retry+1) * 100 * time.Millisecond) // Exponential backoff
}
if success {
atomic.AddInt64(&totalProduced, 1)
successCount++
} else {
atomic.AddInt64(&totalErrors, 1)
if atomic.LoadInt64(&totalErrors) < 10 {
glog.Errorf("Producer %d failed record %d after retries: %v", producerID, recordID, lastErr)
}
}
// Batch progress logging
if successCount > 0 && successCount%10000 == 0 {
glog.Infof("Producer %d: %d/%d records completed", producerID, successCount, recordsPerProducer)
}
// Small delay to prevent overwhelming the broker
if i > 0 && i%batchSize == 0 {
time.Sleep(10 * time.Millisecond)
}
}
glog.Infof("Producer %d completed: %d successful, %d errors",
producerID, successCount, recordsPerProducer-successCount)
return nil
}
// Start concurrent producers
glog.Infof("Starting FIXED %d producers for %d records total", numProducers, totalRecords)
var wg sync.WaitGroup
recordsPerProducer := totalRecords / numProducers
for i := 0; i < numProducers; i++ {
wg.Add(1)
go func(producerID int) {
defer wg.Done()
if err := producer(producerID, recordsPerProducer); err != nil {
glog.Errorf("Producer %d FAILED: %v", producerID, err)
}
}(i)
}
// Wait for completion with timeout
done := make(chan bool)
go func() {
wg.Wait()
done <- true
}()
select {
case <-done:
glog.Infof("All producers completed normally")
case <-time.After(30 * time.Minute): // 30-minute timeout
glog.Errorf("Test timed out after 30 minutes")
t.Errorf("Test timed out")
return
}
produceTime := time.Since(startTime)
finalProduced := atomic.LoadInt64(&totalProduced)
finalErrors := atomic.LoadInt64(&totalErrors)
// Performance results
throughputPerSec := float64(finalProduced) / produceTime.Seconds()
dataVolumeMB := float64(finalProduced) * 300 / (1024 * 1024)
throughputMBPerSec := dataVolumeMB / produceTime.Seconds()
successRate := float64(finalProduced) / float64(totalRecords) * 100
glog.Infof("\n"+
"=== FINAL MILLION RECORD TEST RESULTS ===\n"+
"==========================================\n"+
"Records produced: %d / %d\n"+
"Production time: %v\n"+
"Average throughput: %.0f records/sec\n"+
"Data volume: %.1f MB\n"+
"Bandwidth: %.1f MB/sec\n"+
"Errors: %d (%.2f%%)\n"+
"Success rate: %.1f%%\n"+
"Partitions used: %d\n"+
"Concurrent producers: %d\n",
finalProduced, totalRecords,
produceTime,
throughputPerSec,
dataVolumeMB,
throughputMBPerSec,
finalErrors,
float64(finalErrors)/float64(totalRecords)*100,
successRate,
numPartitions,
numProducers,
)
// Test assertions
if finalProduced < int64(totalRecords*0.95) { // Allow 5% tolerance
t.Errorf("Too few records produced: %d < %d (95%% of target)", finalProduced, int64(float64(totalRecords)*0.95))
}
if finalErrors > int64(totalRecords*0.05) { // Error rate should be < 5%
t.Errorf("Too many errors: %d > %d (5%% of target)", finalErrors, int64(float64(totalRecords)*0.05))
}
if throughputPerSec < 100 {
t.Errorf("Throughput too low: %.0f records/sec (expected > 100)", throughputPerSec)
}
glog.Infof("🏆 MILLION RECORD KAFKA INTEGRATION TEST COMPLETED SUCCESSFULLY!")
}
|
