Flink CDC MySqlSplitReader问题

Tue, 19 Dec 2023 00:45:24 -0800 

我在阅读flink-connector-mysql-cdc项目源码过程中遇到一个不清楚的地方，还请大佬指点，谢谢！


MySqlSplitReader类有一段代码如下，注释“(1) Reads binlog split firstly and then read 
snapshot split”这一句话我不理解。
为什么要先读binlog split再读snapshot split？为保证记录的时序性，不是应该先读全量的snapshot split再读增量的binlog 
split么？


private MySqlRecords pollSplitRecords() throws InterruptedException {
        Iterator<SourceRecords> dataIt;
        if (currentReader == null) {
            // (1) Reads binlog split firstly and then read snapshot split
            if (binlogSplits.size() > 0) {
                // the binlog split may come from:
                // (a) the initial binlog split
                // (b) added back binlog-split in newly added table process
                MySqlSplit nextSplit = binlogSplits.poll();
                currentSplitId = nextSplit.splitId();
                currentReader = getBinlogSplitReader();
                currentReader.submitSplit(nextSplit);
            } else if (snapshotSplits.size() > 0) {
                MySqlSplit nextSplit = snapshotSplits.poll();
                currentSplitId = nextSplit.splitId();
                currentReader = getSnapshotSplitReader();
                currentReader.submitSplit(nextSplit);
            } else {
                LOG.info("No available split to read.");
            }
            dataIt = currentReader.pollSplitRecords();
            return dataIt == null ? finishedSplit() : forRecords(dataIt);
        } else if (currentReader instanceof SnapshotSplitReader) {
          ....
        }
        ...
}

回复