In this deep dive blog post, I will be exploring the critical and often misunderstood dynamics of data loss within replicated systems. Our main focus will be on the potentially catastrophic impact of losing unsynchronized data on a single node that could cause global data loss. In the end, I will illustrate it with a demo.
What Is fsync?
By default, the disk write API is asynchronous. Therefore, when an application uses the operating system’s API to write data to disk, the OS copies the data and may confirm the write request without waiting for the data to reach the disk. This behavior improves latency and throughput, but it reduces safety.