Flink CDC 源码学习(五)

2023-11-15  本文已影响0人  无色的叶

MySqlSourceReader类核心功能

数据分片

执行start方法向MySqlSourceEnumerator发起分片请求, 触发MySqlSourceEnumerator类的handleSplitRequest方法, 并向reader分配分片

 public void start() {
        if (getNumberOfCurrentlyAssignedSplits() <= 1) {
            context.sendSplitRequest();
        }
    }

分片处理

执行addSplits方法, 处理MySqlSourceEnumerator分配的chunk分片

 private void addSplits(List<MySqlSplit> splits, boolean checkTableChangeForBinlogSplit) {
        // restore for finishedUnackedSplits
        // 存放未处理的分片
        List<MySqlSplit> unfinishedSplits = new ArrayList<>();
        for (MySqlSplit split : splits) {
            LOG.info("Source reader {} adds split {}", subtaskId, split);
            // 全量阶段
            if (split.isSnapshotSplit()) {
                MySqlSnapshotSplit snapshotSplit = split.asSnapshotSplit();
                // 判断该分片是否已完成, 如已完成存放到finishedUnackedSplits map中
                if (snapshotSplit.isSnapshotReadFinished()) {
                    finishedUnackedSplits.put(snapshotSplit.splitId(), snapshotSplit);
                } else if (sourceConfig
                        .getTableFilters()
                        .dataCollectionFilter()
                        .isIncluded(split.asSnapshotSplit().getTableId())) {
                    // 如匹配到配置的同步表, 放入未处理分片的列表中
                    unfinishedSplits.add(split);
                } else {
                    LOG.debug(
                            "The subtask {} is skipping split {} because it does not match new table filter.",
                            subtaskId,
                            split.splitId());
                }
            } else {
                // Binlog阶段
                MySqlBinlogSplit binlogSplit = split.asBinlogSplit();
                // When restore from a checkpoint, the finished split infos may contain some splits
                // for the deleted tables.
                // We need to remove these splits for the deleted tables at the finished split
                // infos.
                // 如果从CK恢复, 需移除可能被删除的同步表
                if (checkTableChangeForBinlogSplit) {
                    binlogSplit =
                            filterOutdatedSplitInfos(
                                    binlogSplit,
                                    sourceConfig
                                            .getMySqlConnectorConfig()
                                            .getTableFilters()
                                            .dataCollectionFilter());
                }

                // Try to discovery table schema once for newly added tables when source reader
                // start or restore
                // 是否有新增同步表
                boolean checkNewlyAddedTableSchema =
                        !mySqlSourceReaderContext.isHasAssignedBinlogSplit()
                                && sourceConfig.isScanNewlyAddedTableEnabled();
                mySqlSourceReaderContext.setHasAssignedBinlogSplit(true);

                // the binlog split is suspended
                if (binlogSplit.isSuspended()) {
                    suspendedBinlogSplit = binlogSplit;
                } else if (!binlogSplit.isCompletedSplit()) {
                    uncompletedBinlogSplits.put(binlogSplit.splitId(), binlogSplit);
                    requestBinlogSplitMetaIfNeeded(binlogSplit);
                } else {
                    uncompletedBinlogSplits.remove(binlogSplit.splitId());
                    MySqlBinlogSplit mySqlBinlogSplit =
                            discoverTableSchemasForBinlogSplit(
                                    binlogSplit, sourceConfig, checkNewlyAddedTableSchema);
                    unfinishedSplits.add(mySqlBinlogSplit);
                }
                LOG.info(
                        "Source reader {} received the binlog split : {}.", subtaskId, binlogSplit);
                context.sendSourceEventToCoordinator(new BinlogSplitAssignedEvent());
            }
        }
        // notify split enumerator again about the finished unacked snapshot splits
        reportFinishedSnapshotSplitsIfNeed();
        // add all un-finished splits (including binlog split) to SourceReaderBase
        // 调用父类SourceReaderBase.addSplits方法, 添加 un-finished splits
        if (!unfinishedSplits.isEmpty()) {
            super.addSplits(unfinishedSplits);
        } else if (suspendedBinlogSplit
                != null) { // only request new snapshot split if the binlog split is suspended
            context.sendSplitRequest();
        }
    }

继续调用SourceReaderBase.addSplits方法

 @Override
    public void addSplits(List<SplitT> splits) {
        LOG.info("Adding split(s) to reader: {}", splits);
        // Initialize the state for each split.
        splits.forEach(
                s ->
                        splitStates.put(
                                s.splitId(), new SplitContext<>(s.splitId(), initializedState(s))));
        // Hand over the splits to the split fetcher to start fetch.
        splitFetcherManager.addSplits(splits);
    }

继续调用splitFetcherManager.addSplits方法, 其中splitFetcherManager为SingleThreadFetcherManager对象, 在MySqlSourceReader构造函数中进行的初始化

 @Override
    public void addSplits(List<SplitT> splitsToAdd) {
        // 获取正在运行的fetcher
        SplitFetcher<E, SplitT> fetcher = getRunningFetcher();
        if (fetcher == null) {
            /**
             *  创建SplitFetcher, 并加入到fetchers集合中
             * */
            fetcher = createSplitFetcher();
            // Add the splits to the fetchers.
            // 创建AddSplitsTask任务, 并加入到taskQueue中待执行
            fetcher.addSplits(splitsToAdd);
            // 启动执行fetcher
            startFetcher(fetcher);
        } else {
            // 创建AddSplitsTask任务, 并加入到taskQueue中待执行
            fetcher.addSplits(splitsToAdd);
        }
    }

继续查看SplitFetcher run方法

 @Override
    public void run() {
        LOG.info("Starting split fetcher {}", id);
        try {
            while (runOnce()) {
                // nothing to do, everything is inside #runOnce.
            }
        } catch (Throwable t) {
            errorHandler.accept(t);
        } finally {
            try {
                splitReader.close();
            } catch (Exception e) {
                errorHandler.accept(e);
            } finally {
                LOG.info("Split fetcher {} exited.", id);
                // This executes after possible errorHandler.accept(t). If these operations bear
                // a happens-before relation, then we can checking side effect of
                // errorHandler.accept(t)
                // to know whether it happened after observing side effect of shutdownHook.run().
                shutdownHook.run();
            }
        }
    }

继续调用runOnce()

boolean runOnce() {
        // first blocking call = get next task. blocks only if there are no active splits and queued
        // tasks.
        SplitFetcherTask task;
        lock.lock();
        try {
            if (closed) {
                return false;
            }
            // 从taskQueue获取待执行的任务
            task = getNextTaskUnsafe();
            if (task == null) {
                // (spurious) wakeup, so just repeat
                return true;
            }

            LOG.debug("Prepare to run {}", task);
            // store task for #wakeUp
            this.runningTask = task;
        } finally {
            lock.unlock();
        }

        // execute the task outside of lock, so that it can be woken up
        boolean taskFinished;
        try {
            // 执行SplitFetcherTask run方法
            taskFinished = task.run();
        } catch (Exception e) {
            throw new RuntimeException(
                    String.format(
                            "SplitFetcher thread %d received unexpected exception while polling the records",
                            id),
                    e);
        }

        // re-acquire lock as all post-processing steps, need it
        lock.lock();
        try {
            this.runningTask = null;
            processTaskResultUnsafe(task, taskFinished);
        } finally {
            lock.unlock();
        }
        return true;
    }

继续调用AddSplitsTask run方法, 这里的splitReader对象是MySqlSplitReader, 执行其handleSplitsChanges方法,

 @Override
    public boolean run() {
        for (SplitT s : splitsToAdd) {
            assignedSplits.put(s.splitId(), s);
        }
        splitReader.handleSplitsChanges(new SplitsAddition<>(splitsToAdd));
        return true;
    }

handleSplitsChanges方法只是把MySqlSplit添加到对应的snapshotSplits或者binlogSplits列表中

@Override
    public void handleSplitsChanges(SplitsChange<MySqlSplit> splitsChanges) {
        if (!(splitsChanges instanceof SplitsAddition)) {
            throw new UnsupportedOperationException(
                    String.format(
                            "The SplitChange type of %s is not supported.",
                            splitsChanges.getClass()));
        }

        LOG.info("Handling split change {}", splitsChanges);
        for (MySqlSplit mySqlSplit : splitsChanges.splits()) {
            if (mySqlSplit.isSnapshotSplit()) {
                snapshotSplits.add(mySqlSplit.asSnapshotSplit());
            } else {
                binlogSplits.add(mySqlSplit.asBinlogSplit());
            }
        }
    }

在前面createSplitFetcher方法创建SplitFetcher对象时, 其构造函数中会创建FetchTask

SplitFetcher(
            int id,
            FutureCompletingBlockingQueue<RecordsWithSplitIds<E>> elementsQueue,
            SplitReader<E, SplitT> splitReader,
            Consumer<Throwable> errorHandler,
            Runnable shutdownHook,
            Consumer<Collection<String>> splitFinishedHook,
            boolean allowUnalignedSourceSplits) {
        this.id = id;
        this.elementsQueue = checkNotNull(elementsQueue);
        this.splitReader = checkNotNull(splitReader);
        this.errorHandler = checkNotNull(errorHandler);
        this.shutdownHook = checkNotNull(shutdownHook);
        this.allowUnalignedSourceSplits = allowUnalignedSourceSplits;

        // 创建fetchTask, 其中run 方法中会调用splitReader.fetch()方法
        this.fetchTask =
                new FetchTask<>(
                        splitReader,
                        elementsQueue,
                        ids -> {
                            ids.forEach(assignedSplits::remove);
                            splitFinishedHook.accept(ids);
                            LOG.info("Finished reading from splits {}", ids);
                        },
                        id);
    }

最终调用MySqlSplitReader.fetch()方法去拉取数据, 由DebeziumReader读取数据

上一篇下一篇

猜你喜欢

热点阅读