基本流程

String loadBalancersName = loadBalancers.get(loadBalancerType);
SendLoadBalancer sendLoadBalancer = registerBeanHandler.getBean(loadBalancersName, SendLoadBalancer.class);
String channelLabel = sendLoadBalancer.chooseChannel(smsTemplateVOs,mobiles);;

根据短信模板，选择负载均衡器

根据负载均衡器，选择channel配置文件

根据配置的channel 进行短信发送

负载均衡器

有若干种类

负载均衡器需要实现两个接口

1. 选择channel

2. 获得所有channel

public interface SendLoadBalancer {/**** @description 根据模板、电话负载均衡发送* @param SmsTemplates 模板列表* @param mobile 电话号码* @return 通道类型*/String chooseChannel(List<SmsTemplateVO> SmsTemplates, Set<String> mobile);/**** @description 查询拥有当前模板，活跃状态的通道* @param SmsTemplates 模板列表* @return 通道类型*/Map<String, String> getChannelList(List<SmsTemplateVO> SmsTemplates);
}

channel有两点参数

1. channel        阿里云、腾讯

2. level        权重

@Override
public Map<String, String> getChannelList(List<SmsTemplateVO> SmsTemplates) {
//处理模板
Set<String> channelLabelList = SmsTemplates.stream()
.map(SmsTemplateVO::getChannelLabel).collect(Collectors.toSet());
//查询模板对应的渠道
List<SmsChannelVO> smsChannels =smsChannelService.findChannelInChannelLabel(channelLabelList);
if (!EmptyUtil.isNullOrEmpty(smsChannels)){
return smsChannels.stream()
.collect(Collectors.toMap(SmsChannelVO::getChannelLabel, SmsChannelVO::getLevel));
}
return null;
}

Hash负载均衡

@Component
public class HashSend extends BaseSendLoadBalancer {@Overridepublic String chooseChannel(List<SmsTemplateVO> SmsTemplates, Set<String> mobile) {//获得当前模板对应的渠道Map<String, String> channelMap = super.getChannelList(SmsTemplates);//取得channel地址ListSet<String> keySet = channelMap.keySet();ArrayList<String> keyList = new ArrayList<String>();keyList.addAll(keySet);// 使用mobile取余int hashCode = mobile.hashCode();int serverListSize = keyList.size();int serverPos = hashCode % serverListSize;return keyList.get(serverPos);}}

1. 根据手机号生成hashcode

2. 对channel长度取余

3. 获得余数

随机

    @Overridepublic String chooseChannel(List<SmsTemplateVO> SmsTemplates, Set<String> mobile) {//获得当前模板对应的渠道Map<String, String> channelMap = super.getChannelList(SmsTemplates);// 取得channel地址ListSet<String> keySet = channelMap.keySet();ArrayList<String> keyList = new ArrayList<String>();keyList.addAll(keySet);Random random = new Random();int randomPos = random.nextInt(keyList.size());return keyList.get(randomPos);}

生成个随机数，然后取余

轮询

    @Overridepublic String chooseChannel(List<SmsTemplateVO> SmsTemplates, Set<String> mobile) {//获得当前模板对应的渠道Map<String, String> channelMap = super.getChannelList(SmsTemplates);// 取得通道地址ListSet<String> keySet = channelMap.keySet();ArrayList<String> keyList = new ArrayList<String>();keyList.addAll(keySet);String channelName = null;synchronized (pos) {if (pos >= keySet.size())pos = 0;channelName = keyList.get(pos);pos ++;}return channelName;}

对pos上锁，让pos++

如果大于channel大小，让pos重置

根据pos获得channel

加权随机

    @Overridepublic String chooseChannel(List<SmsTemplateVO> SmsTemplates, Set<String> mobile) {//获得当前模板对应的渠道Map<String, String> channelMap = super.getChannelList(SmsTemplates);// 取得channel地址ListSet<String> keySet = channelMap.keySet();Iterator<String> iterator = keySet.iterator();List<String> serverList = new ArrayList<String>();while (iterator.hasNext()) {String channel = iterator.next();int weight = Integer.valueOf(channelMap.get(channel));for (int i = 0; i < weight; i++)serverList.add(channel);}Random random = new Random();int randomPos = random.nextInt(serverList.size());return serverList.get(randomPos);}

channel * 权重，放入serverList

在生成随机数

加权轮询

 @Overridepublic String chooseChannel(List<SmsTemplateVO> SmsTemplates, Set<String> mobile) {//获得当前模板对应的渠道Map<String, String> channelMap = super.getChannelList(SmsTemplates);// 取得channel地址ListSet<String> keySet = channelMap.keySet();Iterator<String> iterator = keySet.iterator();List<String> channelListHandler = new ArrayList<String>();while (iterator.hasNext()) {String server = iterator.next();Integer weight = Integer.valueOf(channelMap.get(server));for (int i = 0; i < weight; i++)channelListHandler.add(server);}String channelName = null;synchronized (pos) {if (pos > keySet.size())pos = 0;channelName = channelListHandler.get(pos);pos++;}return channelName;}

和加权随机一样