SDK Operation Instructions

Network Type

Currently, only VPCs are supported.

SDK Working Principles

Data consumption and message confirmation are two asynchronous threads started based on the transaction sequence. The two threads are independent of each other and comply with the causal sequence. The received messages invoke the notify function registered by users in sequence. The SDK ensures that each message is pushed only once.

Importing SDK JAR Package

The SDK JAR package can be directly imported only through the library. Currently, the Maven central repository is not supported.

<dependency>
<groupId>com.huawei.hwclouds</groupId>
<artifactId>drs-subscribe-sdk</artifactId>
<version>1.0</version>
</dependency>

Configuring Parameters

During data subscription, some parameters affect the speed and interval for obtaining data.

Before starting the SDK, you can set the following parameters in the subscribe.properties configuration file as required:

#Initial delay for obtaining the subscription data
MESSAGE_DELAY_TIME = 1500000
#Interval for obtaining the subscription data
MESSAGE_PERIOD_TIME = 2000000
#Notify the user that the subscription data initial delay
NOTIFY_DELAY_TIME = 2000
#Notify the user of the subscription data interval
NOTIFY_PERIOD_TIME = 1000
#Initial delay of the ACK messages from the server
ACK_DELAY_TIME = 3000
#Interval between the ACK messages from the server
ACK_PERIOD_TIME = 5000

• The preceding parameter values are set by default. You can also customize parameter values based on the site requirements.
• The unit of the time mentioned above is microsecond.

Confirmation Mechanism

The SDK uses the automatic batch confirmation mechanism. The client program does not need to invoke the confirmation function and can be confirmed repeatedly.

For example: the client receives five batches of messages, and the server only confirms the first, the second, and the fifth batch, respectively. Messages are confirmed in sequence. Therefore, it indicates that the client has consumed all messages in the first to fifth batch. If the client program is suspended, the consumption checkpoint starts from the fifth batch.

SDK Template

import com.huawei.hwclouds.drs.context.SubscribeContext;
import com.huawei.hwclouds.drs.message.ClusterMessage;
import com.huawei.hwclouds.drs.subscribe.ClusterListener;
import com.huawei.hwclouds.drs.subscribe.DefaultSubscribeClient;
import com.huawei.hwclouds.drs.subscribe.SubscribeClient;
import java.util.List;
public class MainClass {
public static void main(String[] args) throws Exception {
SubscribeContext context = new SubscribeContext();
//Set a Username of the current cloud account.
context.setDomainName("username");
//Set the Password of the current cloud account.
context.setPassword("userpassword");
//Set the IP address of the subscription instance on the Data Subscription Management page.
context.setIp("SubscribeChannelIp");
context.setUserId("userId");
SubscribeClient client =
DefaultSubscribeClient.getSubscribeClient(context);
ClusterListener clusterListener = new ClusterListener() {
@Override
//Client consumption behavior defined in notify
public void notify(List<ClusterMessage> var1) throws Exception {
for (ClusterMessage message : var1) {
System.out.println("Message is " + message.toString());
}
}
};
client.addClusterListener(clusterListener);
client.start();
}
}