Go implementation of the Kafka Connect Couchbase.
Go Dcp Kafka streams documents from Couchbase Database Change Protocol (DCP) and publishes Kafka events in near real-time.
- Less resource usage and higher throughput(see Benchmarks).
- Custom Kafka key and headers implementation(see Example).
- Sending multiple Kafka events for a DCP event(see Example).
- Handling different DCP events such as expiration, deletion and mutation(see Example).
- Kafka compression support(Gzip, Snappy, Lz4, Zstd).
- Kafka producer acknowledges support(fire-and-forget, wait for the leader, wait for the full ISR).
- Metadata can be saved to Couchbase or Kafka.
- Managing batch configurations such as maximum batch size, batch bytes, batch ticker durations.
- Scale up and down by custom membership algorithms(Couchbase, KubernetesHa, Kubernetes StatefulSet or Static, see examples).
- Easily manageable configurations.
The benchmark was made with the 1,001,006 Couchbase document, because it is possible to more clearly observe the difference in the batch structure between the two packages. Default configurations for Java Kafka Connect Couchbase used for both connectors.
Package | Time to Process Events | Average CPU Usage(Core) | Average Memory Usage |
---|---|---|---|
Go Dcp Kafka(1.20) | 12s | 0.383 | 428MB |
Java Kafka Connect Couchbase(JDK11) | 19s | 1.5 | 932MB |
func mapper(event couchbase.Event) []message.KafkaMessage {
// return nil if you wish to discard the event
return []message.KafkaMessage{
{
Headers: nil,
Key: event.Key,
Value: event.Value,
},
}
}
func main() {
c, err := dcpkafka.NewConnector(&config.Connector{
Dcp: dcpConfig.Dcp{
Hosts: []string{"localhost:8091"},
Username: "user",
Password: "password",
BucketName: "dcp-test",
Dcp: dcpConfig.ExternalDcp{
Group: dcpConfig.DCPGroup{
Name: "groupName",
Membership: dcpConfig.DCPGroupMembership{
RebalanceDelay: 3 * time.Second,
},
},
},
Metadata: dcpConfig.Metadata{
Config: map[string]string{
"bucket": "checkpoint-bucket-name",
"scope": "_default",
"collection": "_default",
},
Type: "couchbase",
},
Debug: true},
Kafka: config.Kafka{
CollectionTopicMapping: map[string]string{"_default": "topic"},
Brokers: []string{"localhost:9092"},
},
}, mapper)
if err != nil {
panic(err)
}
defer c.Close()
c.Start()
}
Check out on go-dcp
Variable | Type | Required | Default | Description |
---|---|---|---|---|
kafka.collectionTopicMapping |
map[string]string | yes | Defines which Couchbase collection events will be sent to which topic,:warning: If topic information is entered in the mapper, it will OVERWRITE this config. | |
kafka.brokers |
[]string | yes | Broker ip and port information | |
kafka.producerBatchSize |
integer | no | 2000 | Maximum message count for batch, if exceed flush will be triggered. |
kafka.producerBatchBytes |
64 bit integer | no | 10mb | Maximum size(byte) for batch, if exceed flush will be triggered. 10mb is default. |
kafka.producerBatchTimeout |
time.duration | no | 1 nano second | Time limit on how often incomplete message batches will be flushed. |
kafka.producerMaxAttempts |
int | no | math.MaxInt | Limit on how many attempts will be made to deliver a message. |
kafka.producerBatchTickerDuration |
time.Duration | no | 10s | Batch is being flushed automatically at specific time intervals for long waiting messages in batch. |
kafka.readTimeout |
time.Duration | no | 30s | segmentio/kafka-go - Timeout for read operations |
kafka.writeTimeout |
time.Duration | no | 30s | segmentio/kafka-go - Timeout for write operations |
kafka.compression |
integer | no | 0 | Compression can be used if message size is large, CPU usage may be affected. 0=None, 1=Gzip, 2=Snappy, 3=Lz4, 4=Zstd |
kafka.requiredAcks |
integer | no | 1 | segmentio/kafka-go - Number of acknowledges from partition replicas required before receiving a response to a produce request. 0=fire-and-forget, do not wait for acknowledgements from the, 1=wait for the leader to acknowledge the writes, -1=wait for the full ISR to acknowledge the writes |
kafka.secureConnection |
bool | no | false | Enable secure Kafka. |
kafka.rootCAPath |
string | no | *not set | Define root CA path. |
kafka.interCAPath |
string | no | *not set | Define inter CA path. |
kafka.scramUsername |
string | no | *not set | Define scram username. |
kafka.scramPassword |
string | no | *not set | Define scram password. |
kafka.metadataTTL |
time.Duration | no | 60s | TTL for the metadata cached by segmentio, increase it to reduce network requests. For more detail please check docs. |
kafka.metadataTopics |
[]string | no | Topic names for the metadata cached by segmentio, define topics here that the connector may produce. In large Kafka clusters, this will reduce memory usage. For more detail please check docs. | |
kafka.clientID |
string | no | Unique identifier that the transport communicates to the brokers when it sends requests. For more detail please check docs. | |
kafka.allowAutoTopicCreation |
bool | no | false | Create topic if missing. For more detail please check docs. |
Variable | Type | Description |
---|---|---|
metadata.type |
string | Metadata storing types. kafka ,file or couchbase . |
metadata.readOnly |
bool | Set this for debugging state purposes. |
metadata.config |
map[string]string | Set key-values of config. topic ,partition ,replicationFactor for kafka type |
Metric Name | Description | Labels | Value Type |
---|---|---|---|
cbgo_kafka_connector_latency_ms_current | Time to adding to the batch. | N/A | Gauge |
cbgo_kafka_connector_batch_produce_latency_ms_current | Time to produce messages in the batch. | N/A | Gauge |
You can also use all DCP-related metrics explained here. All DCP-related metrics are automatically injected. It means you don't need to do anything.
Date taking effect | Date announced | Change | How to check |
---|---|---|---|
November 11, 2023 | November 11, 2023 | Creating connector via builder | Compile project |
Go Dcp Kafka is always open for direct contributions. For more information please check our Contribution Guideline document.
Released under the MIT License.