Ordered messages, also known as first-in-first-out (FIFO) messages, are provided by Message Queue for Apache RocketMQ. Ordered messages are published and consumed in a strict order. This topic provides sample code to show you how to send and subscribe to ordered messages by using HTTP client SDK for Go.

Background information

Ordered messages are classified into the following types:
  • Globally ordered message: All messages in a specified topic are published and consumed in first-in-first-out (FIFO) order.
  • Partitionally ordered message: All messages in a specified topic are distributed to different partitions by using the partition key. The messages in each partition are published and consumed in FIFO order. A partition key is a key field that is used to distinguish different partitions for ordered messages. A partition key is different from the key of a normal message.

For more information, see Ordered messages.

Prerequisites

The following operations are performed:

  • Install the SDK for Go. For more information, see Prepare the environment.
  • Create resources that you want to specify in the code. For example, you must create the instance, topic, and group that you want to specify in the code in the Message Queue for Apache RocketMQ console in advance. For more information, see Create resources.

Send ordered messages

Notice The Message Queue for Apache RocketMQ broker determines the order in which messages are generated based on the order in which the sender uses a single producer to concurrently send messages in a single thread. If the sender uses multiple producers or multiple threads to concurrently send messages, the message order is determined based on the order in which the messages are received by the Message Queue for Apache RocketMQ broker. This order may be different from the sending order on the business side.

The following sample code shows how to send ordered messages:


package main

import (
    "fmt"
    "time"
    "strconv"

    "github.com/aliyunmq/mq-http-go-sdk"
)

func main() {
    // The HTTP endpoint. To obtain the HTTP endpoint, log on to the Message Queue for Apache RocketMQ console. In the left-side navigation pane, click Instances. On the Instances page, click the name of your instance. On the Instance Details page, scroll to the Basic Information section and view the endpoint on the Endpoints tab. 
    endpoint := "${HTTP_ENDPOINT}"
    // The AccessKey ID that you created in the Resource Access Management (RAM) console for identity authentication. 
    accessKey := "${ACCESS_KEY}"
    // The AccessKey secret that you created in the RAM console for identity authentication. 
    secretKey := "${SECRET_KEY}"
    // The topic of the message. The topic is created in the Message Queue for Apache RocketMQ console. 
    topic := "${TOPIC}"
    // The ID of the instance to which the topic belongs. The instance is created in the Message Queue for Apache RocketMQ console. 
    // If the instance has a namespace, the instance ID must be specified. If the instance does not have a namespace, set the instance ID to null or an empty string. You can check whether your instance has a namespace on the Instance Details page in the Message Queue for Apache RocketMQ console. 
    instanceId := "${INSTANCE_ID}"

    client := mq_http_sdk.NewAliyunMQClient(endpoint, accessKey, secretKey, "")

    mqProducer := client.GetProducer(instanceId, topic)
    // Cyclically send eight messages. 
    for i := 0; i < 8; i++ {
        msg := mq_http_sdk.PublishMessageRequest{
            MessageBody: "hello mq!",         // The message content. 
            MessageTag:  "",                  // The message tag. 
            Properties:  map[string]string{}, // The message properties. 
        }
        // The key of the message. 
        msg.MessageKey = "MessageKey"
        // The custom attributes of the message. 
        msg.Properties["a"] = strconv.Itoa(i)
        // The Sharding Key that is used to distribute ordered messages to a specific partition. The key field can be used to identify different partitions. A Sharding Key is different from a message key. 
        msg.ShardingKey = strconv.Itoa(i % 2)
        ret, err := mqProducer.PublishMessage(msg)

        if err != nil {
            fmt.Println(err)
            return
        } else {
            fmt.Printf("Publish ---->\n\tMessageId:%s, BodyMD5:%s, \n", ret.MessageId, ret.MessageBodyMD5)
        }
        time.Sleep(time.Duration(100) * time.Millisecond)
    }
}

Subscribe to ordered messages

The following sample code shows how to subscribe to ordered messages:

package main

import (
    "fmt"
    "github.com/gogap/errors"
    "strings"
    "time"

    "github.com/aliyunmq/mq-http-go-sdk"
)

func main() {
    // The HTTP endpoint. To obtain the HTTP endpoint, log on to the Message Queue for Apache RocketMQ console. In the left-side navigation pane, click Instances. On the Instances page, click the name of your instance. On the Instance Details page, scroll to the Basic Information section and view the endpoint on the Endpoints tab. 
    endpoint := "${HTTP_ENDPOINT}"
    // The AccessKey ID that you created in the RAM console for identity authentication. 
    accessKey := "${ACCESS_KEY}"
    // The AccessKey secret that you created in the RAM console for identity authentication. 
    secretKey := "${SECRET_KEY}"
    // The topic of the message. The topic is created in the Message Queue for Apache RocketMQ console. 
    topic := "${TOPIC}"
    // The ID of the instance to which the topic belongs. The instance is created in the Message Queue for Apache RocketMQ console. 
    // If the instance has a namespace, the instance ID must be specified. If the instance does not have a namespace, set the instance ID to null or an empty string. You can check whether your instance has a namespace on the Instance Details page in the Message Queue for Apache RocketMQ console. 
    instanceId := "${INSTANCE_ID}"
    // The ID of the group that you created in the Message Queue for Apache RocketMQ console. 
    groupId := "${GROUP_ID}"

    client := mq_http_sdk.NewAliyunMQClient(endpoint, accessKey, secretKey, "")

    mqConsumer := client.GetConsumer(instanceId, topic, groupId, "")

    for {
        endChan := make(chan int)
        respChan := make(chan mq_http_sdk.ConsumeMessageResponse)
        errChan := make(chan error)
        go func() {
            select {
            case resp := <-respChan:
                {
                    // Specify the message consumption logic. 
                    var handles []string
                    fmt.Printf("Consume %d messages---->\n", len(resp.Messages))
                    for _, v := range resp.Messages {
                        handles = append(handles, v.ReceiptHandle)
                        fmt.Printf("\tMessageID: %s, PublishTime: %d, MessageTag: %s\n"+
                            "\tConsumedTimes: %d, FirstConsumeTime: %d, NextConsumeTime: %d\n"+
                            "\tBody: %s\n"+
                            "\tProps: %s\n"+
                            "\tShardingKey: %s\n",
                            v.MessageId, v.PublishTime, v.MessageTag, v.ConsumedTimes,
                            v.FirstConsumeTime, v.NextConsumeTime, v.MessageBody, v.Properties, v.ShardingKey)
                    }

                    // If the broker does not receive an acknowledgment (ACK) for a message from the consumer before the period of time specified by the NextConsumeTime parameter elapses, the message is consumed again. 
                    // A unique timestamp is specified for the handle of a message each time the message is consumed. 
                    ackerr := mqConsumer.AckMessage(handles)
                    if ackerr != nil {
                        // The broker may fail to receive an ACK for a message from the consumer if the handle of the message times out. 
                        fmt.Println(ackerr)
                        if errAckItems, ok := ackerr.(errors.ErrCode).Context()["Detail"].([]mq_http_sdk.ErrAckItem); ok {
                           for _, errAckItem := range errAckItems {
                              fmt.Printf("\tErrorHandle:%s, ErrorCode:%s, ErrorMsg:%s\n",
                                 errAckItem.ErrorHandle, errAckItem.ErrorCode, errAckItem.ErrorMsg)
                           }
                        } else {
                           fmt.Println("ack err =", ackerr)
                        }
                        time.Sleep(time.Duration(3) * time.Second)
                    } else {
                        fmt.Printf("Ack ---->\n\t%s\n", handles)
                    }

                    endChan <- 1
                }
            case err := <-errChan:
                {
                    // No messages in the topic are available for consumption. 
                    if strings.Contains(err.(errors.ErrCode).Error(), "MessageNotExist") {
                        fmt.Println("\nNo new message, continue!")
                    } else {
                        fmt.Println(err)
                        time.Sleep(time.Duration(3) * time.Second)
                    }
                    endChan <- 1
                }
            case <-time.After(35 * time.Second):
                {
                    fmt.Println("Timeout of consumer message ??")
                    endChan <- 1
                }
            }
        }()

                // The consumer may pull partitionally ordered messages from multiple partitions. The consumer consumes the messages in each partition in the order that the messages are sent. 
                // For example, a consumer pulls partitionally ordered messages from a partition. If the broker does not receive the ACK that a message is consumed, the consumer consumes the message again. 
                // The consumer can consume the next batch of messages from a partition only after all messages that are pulled from the partition in the previous batch are acknowledged as consumed. 
                // In long polling mode, the network timeout period is 35 seconds. 
                // In long polling mode, if no message is available for consumption in the topic, the request is suspended on the broker for a specified period of time. If a message is available for consumption during the period, the broker immediately sends a response to the consumer. In this example, the period is set to 3 seconds. 
        mqConsumer.ConsumeMessageOrderly(respChan, errChan,
            3, // The maximum number of messages that can be consumed at a time. In this example, the value is set to 3. The maximum value of this parameter is 16. 
            3, // The duration of a long polling period. Unit: seconds. In this example, the value is set to 3. The maximum value that you can specify is 30. 
        )
        <-endChan
    }
}