What do you mean that the data is actually not stored in the shared buffer. >From the link you provided : "The “shared_buffers” configuration parameter determines how much memory is dedicated to PostgreSQL to use for caching data."
This seems to be the same like other RDBMS. Typically the terminology used is Buffer Pool. BP is used to cache the changes done to a page and the page cleaning process periodically flushes the dirty pages to disk. Checkpoints typically gurantees that at a given time a page in cache is same as on the disk and is usually used as the start time for roll forward recovery. Does PG follow a different architecture ? -----Original Message----- From: Melvin Davidson <melvin6...@gmail.com> NOTE: It is very important you provide the version of PostgreSQL and O/S when addressing this mail list. That is so others searching the archives in the future can determine it's pertenance. That being said, It looks to me like you have a misunderstanding about how PostgreSQL uses shared_buffers. Data is not actually stored in the shared_buffers. Please refer to the following url which explains it's usage. http://leopard.in.ua/2013/09/05/postgresql-sessting-shared-memory -- Melvin Davidson I reserve the right to fantasize. Whether or not you wish to share my fantasy is entirely up to you.