CMU Database Systems - 05 Buffer Pools (Fall 2018)

[@zhiyizhu3040]

Why didn't I discover this earlier!!! Thank you so much for the great material!!!

[@1오징오]

In modern in-memory database, we don't need to consider 'Buffer pools'. right?

[@izharkhankhattak]

*_Nice lecture_*

[@DenisG631]

Great lecture! One question though. With 32 hits in the Postgres example. Does this mean, that no prefetching has been done? With the prefetching, I would assume that it aggressively drops the already read pages (since they are safe to drop) and replace them with the new prefetched pages. 32/44k ratio sounds really bad

[@MrEternalFool]

32 is 32 pages which is 32 * 4k = 128k which is the size of the buffer pool. I think this means no pre-fetching was done. I don't know why.

[@leonardovincent859]

noice!!!