我试图了解如何解释Flink应用程序显示的检查点指标并进行调整。在Flink UI上，对于其中一个运算符（窗口运算符），在Checkpoints选项卡中的Buffered during Alignment列中看到一个大约100 MB的数字。从文档中，我看到
The amount of data buffered during alignments. For exactly-once semantics, Flink aligns the streams at operators that receive multiple input streams, buffering some data for that alignment. The buffered data volume is ideally low - higher amounts means that checkpoint barriers are received at very different times from the different input streams.
checkpoint barriers received at very different times mean是什么？ Buffered during alignment数字似乎只是在检查点之上增长，我认为它正在影响工作绩效。我该如何调整？