motan源码分析五：cluster相关
上一章我们分析了客户端调用服务端相关的源码，但是到了cluster里面的部分我们就没有分析了，本章将深入分析cluster和它的相关支持类。
1.clustersupport的创建过程，上一章的ReferConfig的initRef()方法中调用了相关的创建代码：
for(Iterator iterator = protocols.iterator(); iterator.hasNext();) { ProtocolConfig protocol = (ProtocolConfig)iterator.next(); LoggerUtil.info((new StringBuilder("ProtocolConfig's")).append(protocol.getName()).toString()); Map params = new HashMap(); params.put(URLParamType.nodeType.getName(), "referer"); params.put(URLParamType.version.getName(), URLParamType.version.getValue()); params.put(URLParamType.refreshTimestamp.getName(), String.valueOf(System.currentTimeMillis())); collectConfigParams(params, new AbstractConfig[] { protocol, basicReferer, extConfig, this }); collectMethodConfigParams(params, getMethods()); URL refUrl = new URL(protocol.getName(), localIp, 0, interfaceClass.getName(), params); ClusterSupport clusterSupport = createClusterSupport(refUrl, configHandler, registryUrls);//创建clustersupport clusterSupports.add(clusterSupport); clusters.add(clusterSupport.getCluster());//获取对应的cluster proxy = proxy != null ? proxy : refUrl.getParameter(URLParamType.proxy.getName(), URLParamType.proxy.getValue()); } private ClusterSupport createClusterSupport(URL refUrl, ConfigHandler configHandler, List registryUrls) { List regUrls = new ArrayList(); if(StringUtils.isNotBlank(directUrl) || "injvm".equals(refUrl.getProtocol())) { URL regUrl = new URL("local", "127.0.0.1", 0, com/weibo/api/motan/registry/RegistryService.getName()); if(StringUtils.isNotBlank(directUrl)) { StringBuilder duBuf = new StringBuilder(128); String dus[] = MotanConstants.COMMA_SPLIT_PATTERN.split(directUrl); String as[]; int j = (as = dus).length; for(int i = 0; i < j; i++) { String du = as[i]; if(du.contains(":")) { String hostPort[] = du.split(":"); URL durl = refUrl.createCopy(); durl.setHost(hostPort[0].trim()); durl.setPort(Integer.parseInt(hostPort[1].trim())); durl.addParameter(URLParamType.nodeType.getName(), "service"); duBuf.append(StringTools.urlEncode(durl.toFullStr())).append(","); } } if(duBuf.length() > 0) { duBuf.deleteCharAt(duBuf.length() - 1); regUrl.addParameter(URLParamType.directUrl.getName(), duBuf.toString()); } } regUrls.add(regUrl); } else//走注册中心的方式 { if(registryUrls == null || registryUrls.isEmpty()) throw new IllegalStateException(String.format("No registry to reference %s on the consumer %s , please config <motan:registry address=\"...\" /> in your spring config.", new Object[] { interfaceClass, "127.0.0.1" })); URL url; for(Iterator iterator = registryUrls.iterator(); iterator.hasNext(); regUrls.add(url.createCopy())) url = (URL)iterator.next(); } URL url; for(Iterator iterator1 = regUrls.iterator(); iterator1.hasNext(); url.addParameter(URLParamType.embed.getName(), StringTools.urlEncode(refUrl.toFullStr()))) url = (URL)iterator1.next(); return configHandler.buildClusterSupport(interfaceClass, regUrls);//调用simpleconfighandler的创建clustersupport方法 } public <T> ClusterSupport<T> buildClusterSupport(Class<T> interfaceClass, List<URL> registryUrls) { ClusterSupport<T> clusterSupport = new ClusterSupport<T>(interfaceClass, registryUrls);//创建cluster支持类，将业务接口和注册中心信息传递进去 clusterSupport.init();//初始化 return clusterSupport; }
2.clustersupport的init和prepare方法
public void init() {
prepareCluster();
URL subUrl = toSubscribeUrl(url);
for (URL ru : registryUrls) {//循环注册中心的url
String directUrlStr = ru.getParameter(URLParamType.directUrl.getName());
// 如果有directUrl，直接使用这些directUrls进行初始化，不用到注册中心discover
if (StringUtils.isNotBlank(directUrlStr)) {
List<URL> directUrls = parseDirectUrls(directUrlStr);
if (!directUrls.isEmpty()) {
notify(ru, directUrls);
LoggerUtil.info("Use direct urls, refUrl={}, directUrls={}", url, directUrls);
continue;
}
}
// client 注册自己，同时订阅service列表
Registry registry = getRegistry(ru);//获取zookeeper的注册中心
registry.subscribe(subUrl, this);//注册自己并订阅服务
}
boolean check = Boolean.parseBoolean(url.getParameter(URLParamType.check.getName(), URLParamType.check.getValue()));
if (!CollectionUtil.isEmpty(cluster.getReferers()) || !check) {
cluster.init();//初始化集群
if (CollectionUtil.isEmpty(cluster.getReferers()) && !check) {
LoggerUtil.warn(String.format("refer:%s", this.url.getPath() + "/" + this.url.getVersion()), "No services");
}
return;
}
throw new MotanFrameworkException(String.format("ClusterSupport No service urls for the refer:%s, registries:%s",
this.url.getIdentity(), registryUrls), MotanErrorMsgConstant.SERVICE_UNFOUND);
}
private void prepareCluster() {
String clusterName = url.getParameter(URLParamType.cluster.getName(), URLParamType.cluster.getValue());//集群名称
String loadbalanceName = url.getParameter(URLParamType.loadbalance.getName(), URLParamType.loadbalance.getValue());//负载均衡名称
String haStrategyName = url.getParameter(URLParamType.haStrategy.getName(), URLParamType.haStrategy.getValue());//ha高可用名称
cluster = ExtensionLoader.getExtensionLoader(Cluster.class).getExtension(clusterName);//获取具体的集群对象
LoadBalance<T> loadBalance = ExtensionLoader.getExtensionLoader(LoadBalance.class).getExtension(loadbalanceName);//获取具体的负载均衡方式，目前motan支持6种负载方式
HaStrategy<T> ha = ExtensionLoader.getExtensionLoader(HaStrategy.class).getExtension(haStrategyName);//获取高可用的方式，目前支持两种failfast和failover方式
cluster.setLoadBalance(loadBalance);
cluster.setHaStrategy(ha);
cluster.setUrl(url);
}
3.负载均衡，motan支持6种方式，分别是：轮训、随机、hash、本地服务优先、权重可配置、低并发优先，具体代码可见com.weibo.api.motan.cluster.loadbalance目录，本文我们主要看一下轮训的方式：
public class RoundRobinLoadBalance<T> extends AbstractLoadBalance<T> {
private AtomicInteger idx = new AtomicInteger(0);
@Override
protected Referer<T> doSelect(Request request) {
List<Referer<T>> referers = getReferers();//获取所有服务器的引用
int index = idx.incrementAndGet();//自增
for (int i = 0; i < referers.size(); i++) {
Referer<T> ref = referers.get((i + index) % referers.size());//利用自增数去模，达到轮训的目的
if (ref.isAvailable()) {
return ref;
}
}
return null;
}
@Override
protected void doSelectToHolder(Request request, List<Referer<T>> refersHolder) {
List<Referer<T>> referers = getReferers();
int index = idx.incrementAndGet();
for (int i = 0; i < referers.size(); i++) {
Referer<T> referer = referers.get((i + index) % referers.size());
if (referer.isAvailable()) {
refersHolder.add(referer);
}
}
}
}
4.motan支持failfast和failover两种方式，failfast只调用一次，如果失败则直接返回失败，failover循环调用若干次，直到成功或循环结束后
public Response call(Request request, LoadBalance<T> loadBalance) {
List<Referer<T>> referers = selectReferers(request, loadBalance);//获取所有的引用
if (referers.isEmpty()) {
throw new MotanServiceException(String.format("FailoverHaStrategy No referers for request:%s, loadbalance:%s", request,
loadBalance));
}
URL refUrl = referers.get(0).getUrl();
// 先使用method的配置
int tryCount =
refUrl.getMethodParameter(request.getMethodName(), request.getParamtersDesc(), URLParamType.retries.getName(),
URLParamType.retries.getIntValue());//获取重试次数
// 如果有问题，则设置为不重试
if (tryCount < 0) {
tryCount = 0;
}
for (int i = 0; i <= tryCount; i++) {
Referer<T> refer = referers.get(i % referers.size());//循环调用
try {
request.setRetries(i);
return refer.call(request);
} catch (RuntimeException e) {
// 对于业务异常，直接抛出
if (ExceptionUtil.isBizException(e)) {
throw e;//业务异常退出调用
} else if (i >= tryCount) {
throw e;
}
LoggerUtil.warn(String.format("FailoverHaStrategy Call false for request:%s error=%s", request, e.getMessage()));
}
}
throw new MotanFrameworkException("FailoverHaStrategy.call should not come here!");
}
本章知识点总结：
1.一个cluster有一个cluster的支持类，有一个ha，有一个loadbalance;
2.motan支持6种负载均衡方式；
3.motan支持failover的ha方式；
