The database server uses pages of virtual memory for the data structures needed to implement cursors. These pages are kept
locked in memory between fetch requests so they are readily available when the next fetch request arrives.
To prevent these pages from occupying too much of the cache in low memory environments, a limit is placed on the percentage
of the cache allowed to be used for pinning cursors. You can use the pinned_cursor_percent_of_cache option to adjust this
limit.
The option value is specified as a percentage from 0 to 100, with a default of 10. Setting the option to 0 means that cursor
pages are not pinned between fetch requests.