没有合适的资源？快使用搜索试试~ 我知道了~

首页Pro-Java-Clustering-and-Scalability.pdf

Pro-Java-Clustering-and-Scalability.pdf

Java

cluster

需积分: 9 81 浏览量更新于2023-05-30 评论收藏 3.71MB PDF 举报

身份认证购VIP最低享 7 折!

领优惠券(最高得80元）

试读

142页

My name is Jorge Acetozi, and I’m a Brazilian software engineer who has worked for many years as a Java developer. During my career, I have been interested in subjects such as these: • Linux • Distributed systems • Testing automation • Continuous integration • Continuous delivery • Cloud computing • Virtualization • Containerization • Security Why the varied interests? I just didn’t feel that coding in Java only was enough for me professionally (although doing this while following best practices is not an easy task). I wanted to understand the entire process of creating software and delivering it to a production environment. So, some years ago I started a career as a DevOps engineer. After taking these two paths, I’ve noticed there are two types of software engineer. In the first group are developers who usually don’t feel excited by infrastructure subjects and merely want to write code following best practices. However, this means they are not able to maintain a production environment since it involves much more than just writing software code. In the second group are infrastructure people who usually hate writing software code (note that writing small scripts to automate infrastructure tasks are quite different than writing software code). On the other hand, these people are able to maintain a production environment because they understand the deployment process, how to monitor the servers, how to handle security issues, and so on. The software engineer I’m trying to become sits right in the middle of these types of developers and infrastructure folks. I’d like to be an excellent programmer who follows coding best practices, but I also want to be able to put code into production and maintain it.

资源详情

资源评论

资源推荐

J. Acetozi, Pro Java Clustering and Scalability, DOI 10.1007/978-1-4842-2985-9_18

CHAPTER 18

Joining the Chat Room

When you join a chat room, some JavaScript code gets executed on the client side.

function connect() {

socket = new SockJS('/ws');

stompClient = Stomp.over(socket);

stompClient.connect({ 'chatRoomId' : chatRoomId }, stompSuccess,

stompFailure);

}

function stompSuccess(frame) {

enableInputMessage();

successMessage("Your WebSocket connection was successfully

established!")

stompClient.subscribe('/chatroom/connected.users',

updateConnectedUsers);

stompClient.subscribe('/chatroom/old.messages', oldMessages);

stompClient.subscribe('/topic/' + chatRoomId + '.public.messages',

publicMessages);

stompClient.subscribe('/user/queue/' + chatRoomId + '.private.

messages', privateMessages);

stompClient.subscribe('/topic/' + chatRoomId + '.connected.users',

updateConnectedUsers);

}

function stompFailure(error) {

errorMessage("Lost connection to WebSocket! Reconnecting in

10 seconds...");

disableInputMessage();

setTimeout(connect, 10000);

}

CHAPTER 18 ■ JOINING THE CHAT ROOM

100

First, the connect function is called, and a new WebSocket connection

is started using STOMP as a subprotocol. Note that you also add the header

'chatRoomId' : chatRoomId when you send the CONNECT frame because the

server must keep this information to ensure that it’s manipulating the correct

destination names.

If the handshake succeeds, then the stompSuccess function is called, and

the user subscribes to some destinations. Actually, the first two subscriptions are

executed only once on the server side, and they are used to get some initial data,

that is, his old conversation and the current connected users to this specific chat

room. This is the Java code executed when these two subscriptions happen:

@Controller

public class ChatRoomController {

@Autowired

private ChatRoomService chatRoomService;

@Autowired

private InstantMessageService instantMessageService;

@SubscribeMapping("/connected.users")

public List<ChatRoomUser> listChatRoomConnectedUsersOnSubscribe

(SimpMessageHeaderAccessor headerAccessor) {

String chatRoomId = headerAccessor.getSessionAttributes().get

("chatRoomId").toString();

return chatRoomService.findById(chatRoomId).getConnectedUsers();

}

@SubscribeMapping("/old.messages")

public List<InstantMessage> listOldMessagesFromUserOnSubscribe

(Principal principal, SimpMessageHeaderAccessor headerAccessor) {

String chatRoomId = headerAccessor.getSessionAttributes().get

("chatRoomId").toString();

return instantMessageService.findAllInstantMessagesFor

(principal.getName(), chatRoomId);

}

...

}

CHAPTER 18 ■ JOINING THE CHAT ROOM

101

The List<ChatRoomUser> list is fetched from Redis, and the

List<InstantMessage> list is fetched from Cassandra.

The @SubscribeMapping annotation is useful when users need to fetch the

initial data using the WebSocket connection.

The last three subscriptions in the JavaScript code are the following,

respectively:

• Execute the publicMessages function when a new message

arrives at the '/topic/' + chatRoomId + '.public.

messages' destination. This function will render a public

message in the user’s message panel.

• Execute the privateMessages function when a new

message arrives at the '/user/queue/' + chatRoomId

+ '.private.messages' destination. This function will

render a private message in the user’s message panel.

• Execute the updateConnectedUsers function when a

new message arrives at the '/topic/' + chatRoomId +

'.connected.users' destination. This function will update

the connected users in the user’s connected users panel.

18.1 WebSocket Reconnection Strategy

When the WebSocket connection is lost, the stompFailure function is executed,

and it will try to reestablish the connection every ten seconds. If it succeeds, then

everything explained earlier happens again. Because every message is stored in

Cassandra (regardless of being delivered through the WebSocket connection),

even if the user was offline while someone sent a message to him, when the user

reconnects, the message will be in the List<InstantMessage> list and will be

displayed in the user’s message panel.

18.2 WebSocket Events

Once the WebSocket connection is established or disconnected, an event is

triggered on the server side, and a system message is sent to every connected user

in the chat room informing them that someone has joined or left (Figure 18-1).

CHAPTER 18 ■ JOINING THE CHAT ROOM

102

The following is the code that handles these events on the server side:

@Component

public class WebSocketEvents {

@Autowired

private ChatRoomService chatRoomService;

@EventListener

private void handleSessionConnected(SessionConnectEvent event) {

SimpMessageHeaderAccessor headers = SimpMessageHeaderAccessor.

wrap(event.getMessage());

String chatRoomId = headers.getNativeHeader("chatRoomId").

get(0);

headers.getSessionAttributes().put("chatRoomId", chatRoomId);

ChatRoomUser joiningUser = new ChatRoomUser(event.getUser().

getName());

chatRoomService.join(joiningUser, chatRoomService.

findById(chatRoomId));

}

Figure 18-1. System messages from admin

CHAPTER 18 ■ JOINING THE CHAT ROOM

103

@EventListener

private void handleSessionDisconnect(SessionDisconnectEvent event) {

SimpMessageHeaderAccessor headers = SimpMessageHeaderAccessor.

wrap(event.getMessage());

String chatRoomId = headers.getSessionAttributes().

get("chatRoomId").toString();

ChatRoomUser leavingUser = new ChatRoomUser(event.getUser().

getName());

chatRoomService.leave(leavingUser, chatRoomService.

findById(chatRoomId));

}

Let’s take the SessionConnected event as an example and follow the entire

flow.

The handleSessionConnected method is called when a WebSocket connection

is created. Then, the chatRoomId value is obtained from the CONNECT frame headers,

and it’s stored in the user’s WebSocket session as an attribute. This is convenient

since from now every message the client sends doesn’t need to provide the

chatRoomId value, as it’s already stored in the user’s WebSocket session. Then, the

join method is called, passing joiningUser and chatRoom as parameters.

@Service

public class RedisChatRoomService implements ChatRoomService {

@Autowired

private SimpMessagingTemplate webSocketMessagingTemplate;

@Autowired

private ChatRoomRepository chatRoomRepository;

@Autowired

private InstantMessageService instantMessageService;

@Override

public ChatRoom join(ChatRoomUser joiningUser, ChatRoom chatRoom) {

chatRoom.addUser(joiningUser);

chatRoomRepository.save(chatRoom);

sendPublicMessage(SystemMessages.welcome(chatRoom.getId(),

joiningUser.getUsername()));

updateConnectedUsersViaWebSocket(chatRoom);

return chatRoom;

}

剩余141页未读，继续阅读

可以使用k-means++算法，该算法可以更好地初始化聚类中心，从而提高聚类效果。在聚类时，可以将配送点的时间属性和配送量属性作为特征，然后使用k-means++算法进行聚类。在聚类过程中，可以设置时间属性和配送量属性的权重，以达到相对均衡的聚类效果。具体JAVA代码如何实现？

可以使用以下代码实现 k-means 算法的聚类： ``` import java.util.ArrayList; import java.util.List; import java.util.Random; public class KMeans { private int k; // 聚类数 private int maxIter; // 最大迭代次数 priva

python k-means聚类分析

k-means是一种聚类分析方法，它将数据集划分为k个不重叠的簇。算法流程如下: 1. 随机选择k个数据点作为初始质心 2. 将所有数据点分配到最近的质心所在的簇 3. 对于每个簇，计算所有点的平均值并将其作为新的质心 4. 重复2和3，直到质心不再发生变化。 python有许多第三方库可以实现k-means聚类分析,如scikit-learn,numpy

Intelligent Data Mining - D. Ruan, et al., (Springer, 2005)

目录： Part I: Intelligent Systems and Data Mining Some Considerations in Multi-Source Data Fusion R.R. Yager yager@panix.com Granular Nested Causal Complexes L.J. Mazlack mazlack@uc.edu Gene Regulating

Adventure1995

粉丝: 7
资源: 83

上传资源快速赚钱

我的内容管理收起

我的资源快来上传第一个资源

我的收益

登录查看自己的收益

我的积分登录查看自己的积分

我的C币登录后查看C币余额

我的收藏

我的下载

下载帮助

会员权益专享

Pro-Java-Clustering-and-Scalability.pdf

评论0

会员权益专享

最新资源

Pro-Java-Clustering-and-Scalability.pdf

评论0

用案例说明k-means聚类和层次聚类的区别

K-MEANS(K均值聚类算法,C均值算法)

qiime2 16s双端数据分析代码

Pearson-Digital Image Processing， 4th.Edition-2018.pdf

写一个用k-means算法进行图像聚类的matlab版本代码

可以帮我找一个k-means算法文本聚类的项目案例吗

High Performance in-memory computing with Apache Ignite.pdf

分级聚类有哪些方法可以实现

找几篇PSO优化k_means相关的论文

有什么和k-means类似的聚类算法吗

改进上上段代码中的k-means算法若初始聚类中心存在离群或选择多个距离接近点，则聚类效果不佳的问题

k-means++聚类算法matlab

k-means聚类算法java实现

给我推荐20个比较流行的C++人声分离模型

自行下载公开数据集实现至少4个聚类算法

python k-means聚类分析

Intelligent Data Mining - D. Ruan, et al., (Springer, 2005)

会员权益专享

最新资源