Compatibility with Kafka consumers and usage notes

Franz-go parameters

parameter	Description
kgo.SeedBrokers	The cluster address for initial connection, in the format of Project.Endpoint:Port. Configure this parameter based on the endpoint of the project. For more information, see View endpoints. - Alibaba Cloud internal network: The port number is 10011. Example: project.cn-hangzhou-intranet.log.aliyuncs.com:10011 - Internet: The port number is 10012. Example: project.cn-hangzhou.log.aliyuncs.com:10012
kgo.ConsumerGroup	The ID of the consumer group. You can add consumers to a consumer group. Then, you can specify the ID of the consumer group to implement load balancing, data processing, and data delivery. Example:"kafka-test"
kgo.ConsumeTopics	The name of the Simple Log Service Logstore.
kgo.SASL	To ensure the data transmission security, set the value to SASL_SSL
User	The name of the Simple Log Service Project.
Pass	The AccessKey pair of your Alibaba Cloud account, in the {access-key-id}#{access-key-secret} format. Replace {access-key-id} with your AccessKey ID and {access-key-secret} with your AccessKey secret. We recommend that you use the AccessKey pair of a Resource Access Management (RAM) user. For more information, see Create a RAM user and authorize the RAM user to access Simple Log Service
kgo.Dialer	If you use the Simple Authentication Security Layer (SASL) protocol, you must configure this parameter. Otherwise, consumption fails. For more information, see franz-go on GitHub

Franz-go sample code

package main

import (
	"context"
	"crypto/tls"
	"fmt"
        "os"
	"github.com/twmb/franz-go/pkg/kgo"
	"github.com/twmb/franz-go/pkg/sasl/plain"
	"net"
	"time"
)

func main() {
	project         := "project"
	logstore        := "logstore"
	endpoint        := "cn-hangzhou.log.aliyuncs.com"
	port            := "10012"
	
	groupId         := "kafka-test"
	accessKeyID     := os.Getenv("SLS_ACCESS_KEY_ID")
        accessKeySecret := os.Getenv("SLS_ACCESS_KEY_SECRET")

 	//using sasl we must add tisDialer
	tlsDialer := &tls.Dialer{NetDialer: &net.Dialer{Timeout: 10 * time.Second}}
	seeds := []string{fmt.Sprintf("%s.%s:%s", project, endpoint, port)}

	//get Kgo client
	client, err := kgo.NewClient(
		kgo.SeedBrokers(seeds...),
		kgo.ConsumerGroup(groupId),
		kgo.ConsumeTopics(logstore),
		kgo.SASL(plain.Auth{
			User: project,
			Pass: fmt.Sprintf("%s#%s", accessKeyID, accessKeySecret),
		}.AsMechanism()),
		kgo.Dialer(tlsDialer.DialContext),
	)

	if err != nil {
		panic(err)
	}
	defer client.Close()
	ctx := context.Background()

	for {
		fetches := client.PollFetches(ctx)
		if errs := fetches.Errors(); len(errs) > 0 {
			panic(fmt.Sprint(errs))
		}

		iter := fetches.RecordIter()
		for !iter.Done() {
			record := iter.Next()
			fmt.Println(string(record.Value), "from an iterator!")
		}
	}
}

Comparison between franz-go and confluent-kafka-go

franz-go has 857 stars on GitHub, whereas confluent-kafka-go has 3,700 stars on GitHub. In real applications, franz-go reports fewer errors than confluent-kafka-go in case of a parameter configuration error, or even does not report errors. This increases the troubleshooting difficulty. We recommend that you use confluent-kafka-go.