nacos RaftCore中MasterElection的原理及應用

本篇內(nèi)容主要講解“nacos RaftCore中MasterElection的原理及應用”，感興趣的朋友不妨來看看。本文介紹的方法操作簡單快捷，實用性強。下面就讓小編來帶大家學習“nacos RaftCore中MasterElection的原理及應用”吧!

序

本文主要研究一下nacos RaftCore的MasterElection

RaftCore

nacos-1.1.3/naming/src/main/java/com/alibaba/nacos/naming/consistency/persistent/raft/RaftCore.java

@Component
public class RaftCore {
	
	//......

    @PostConstruct
    public void init() throws Exception {

        Loggers.RAFT.info("initializing Raft sub-system");

        executor.submit(notifier);

        long start = System.currentTimeMillis();

        raftStore.loadDatums(notifier, datums);

        setTerm(NumberUtils.toLong(raftStore.loadMeta().getProperty("term"), 0L));

        Loggers.RAFT.info("cache loaded, datum count: {}, current term: {}", datums.size(), peers.getTerm());

        while (true) {
            if (notifier.tasks.size() <= 0) {
                break;
            }
            Thread.sleep(1000L);
        }

        initialized = true;

        Loggers.RAFT.info("finish to load data from disk, cost: {} ms.", (System.currentTimeMillis() - start));

        GlobalExecutor.registerMasterElection(new MasterElection());
        GlobalExecutor.registerHeartbeat(new HeartBeat());

        Loggers.RAFT.info("timer started: leader timeout ms: {}, heart-beat timeout ms: {}",
            GlobalExecutor.LEADER_TIMEOUT_MS, GlobalExecutor.HEARTBEAT_INTERVAL_MS);
    }

	//......
}

• RaftCore的init方法通過GlobalExecutor.registerMasterElection(new MasterElection())注冊了MasterElection

GlobalExecutor

nacos-1.1.3/naming/src/main/java/com/alibaba/nacos/naming/misc/GlobalExecutor.java

public class GlobalExecutor {

	//......

    public static final long TICK_PERIOD_MS = TimeUnit.MILLISECONDS.toMillis(500L);

    public static void registerMasterElection(Runnable runnable) {
        executorService.scheduleAtFixedRate(runnable, 0, TICK_PERIOD_MS, TimeUnit.MILLISECONDS);
    }

    //......
}

• registerMasterElection方法每隔TICK_PERIOD_MS毫秒調(diào)度一次runnable

MasterElection

nacos-1.1.3/naming/src/main/java/com/alibaba/nacos/naming/consistency/persistent/raft/RaftCore.java

    public class MasterElection implements Runnable {
        @Override
        public void run() {
            try {

                if (!peers.isReady()) {
                    return;
                }

                RaftPeer local = peers.local();
                local.leaderDueMs -= GlobalExecutor.TICK_PERIOD_MS;

                if (local.leaderDueMs > 0) {
                    return;
                }

                // reset timeout
                local.resetLeaderDue();
                local.resetHeartbeatDue();

                sendVote();
            } catch (Exception e) {
                Loggers.RAFT.warn("[RAFT] error while master election {}", e);
            }

        }

        public void sendVote() {

            RaftPeer local = peers.get(NetUtils.localServer());
            Loggers.RAFT.info("leader timeout, start voting,leader: {}, term: {}",
                JSON.toJSONString(getLeader()), local.term);

            peers.reset();

            local.term.incrementAndGet();
            local.voteFor = local.ip;
            local.state = RaftPeer.State.CANDIDATE;

            Map<String, String> params = new HashMap<>(1);
            params.put("vote", JSON.toJSONString(local));
            for (final String server : peers.allServersWithoutMySelf()) {
                final String url = buildURL(server, API_VOTE);
                try {
                    HttpClient.asyncHttpPost(url, null, params, new AsyncCompletionHandler<Integer>() {
                        @Override
                        public Integer onCompleted(Response response) throws Exception {
                            if (response.getStatusCode() != HttpURLConnection.HTTP_OK) {
                                Loggers.RAFT.error("NACOS-RAFT vote failed: {}, url: {}", response.getResponseBody(), url);
                                return 1;
                            }

                            RaftPeer peer = JSON.parseObject(response.getResponseBody(), RaftPeer.class);

                            Loggers.RAFT.info("received approve from peer: {}", JSON.toJSONString(peer));

                            peers.decideLeader(peer);

                            return 0;
                        }
                    });
                } catch (Exception e) {
                    Loggers.RAFT.warn("error while sending vote to server: {}", server);
                }
            }
        }
    }

• MasterElection實現(xiàn)了Runnable方法，其run方法在peers都是ready而且local.leaderDueMs減去TICK_PERIOD_MS小于等于0的時候會開始選舉；它首先resetLeaderDue及resetHeartbeatDue，然后執(zhí)行sendVote方法；sendVote方法首先重置peers，遞增localPeer的term，并設置voteFor為自己，然后更新state為RaftPeer.State.CANDIDATE，最后遍歷peers.allServersWithoutMySelf()，將自己的vote信息異步post給其他peer；如果其他peer返回成功則執(zhí)行peers.decideLeader(peer)，返回1，否則返回0

小結(jié)

RaftCore的init方法通過GlobalExecutor.registerMasterElection(new MasterElection())注冊了MasterElection；registerMasterElection方法每隔TICK_PERIOD_MS毫秒調(diào)度一次；MasterElection實現(xiàn)了Runnable方法，其run方法在peers都是ready而且local.leaderDueMs減去TICK_PERIOD_MS小于等于0的時候會開始選舉；它首先resetLeaderDue及resetHeartbeatDue，然后執(zhí)行sendVote方法

到此，相信大家對“nacos RaftCore中MasterElection的原理及應用”有了更深的了解，不妨來實際操作一番吧！這里是億速云網(wǎng)站，更多相關內(nèi)容可以進入相關頻道進行查詢，關注我們，繼續(xù)學習！