I've often wondered if a more sparse representation of modified data that
*only* contains the modifications would save memory and also give
acceptable lookup performance. Presumably after modification its likely
that a commit will follow shortly, so then it might make sense to penalize
lookups that happen between modifications and commits. Just a thought.

--JYL

> for clarification:
> MK saves modified columns of data on the heap.  when you commit, it
> writes them to the file (via normal i/o or memory mapping, as
> appropriate), and then frees that heap space.   so if the file is mapped
> (as it will be if it is not huge) the total space in use stays about the
> same over a commit.  the virtual space will likely grow, for the first
> commit, as the memory allocator likely keeps the freed heap in the
> virual space.
>
> --- Jacob Levy <[EMAIL PROTECTED]> wrote:
>> Bernard
>>
>> Committing does not free the memory. As I said before, Metakit maps
>> the *entire file* into memory. So if the file grows, so does the
>> memory mapping requirement. Whether the physical RAM used by the
>> program grows or not is a factor influenced by the OS, not by Metakit.
>>
>> --JYL
>
>
> __________________________________
> Do you Yahoo!?
> Yahoo! SiteBuilder - Free web site building tool. Try it!
> http://webhosting.yahoo.com/ps/sb/
> _____________________________________________
> Metakit mailing list  -  [EMAIL PROTECTED]
> http://www.equi4.com/mailman/listinfo/metakit



_____________________________________________
Metakit mailing list  -  [EMAIL PROTECTED]
http://www.equi4.com/mailman/listinfo/metakit

Reply via email to