Hi

I wonder how Oracle manage internally the buffer cache in following situation

The buffer cache is small, so small that if someone do a

select * from table_name


table which has 100 million rows all the data wont fit in the buffer cache, how Oracle manage this? Let´s put some figures, the buffer cache can store data blocks of 10 million rows of this table, will Oracle first return 10 million rows then remove them from the buffer cache then read from disk and put another 10 million rows into buffer cache and remove again etc until all 100 million rows have been fetched and returned to user? i.e 10 fetches instead of one

[Edited by pando on 01-31-2001 at 08:30 AM]