On Thursday, February 3, 2022, James Pang (chaolpan) <chaolpan@cisco.com> wrote:
When Postgres checkpoint, it always write buffers from pg shared_buffers to OS filesystem cache, after that, it finally make a system call fsync , that will flush dirty blocks from OS filesystem cache to disk, right?
Please don’t top-post.
From the docs:
On Linux and POSIX platforms checkpoint_flush_after allows to force the OS that pages written by the checkpoint should be flushed to disk after a configurable number of bytes. Otherwise, these pages may be kept in the OS's page cache, inducing a stall when fsync is issued at the end of a checkpoint. This setting will often help to reduce transaction latency, but it also can have an adverse effect on performance; particularly for workloads that are bigger than shared_buffers, but smaller than the OS's page cache.
