On Tue, 27 Apr 1999 02:38:02 +0200 (CEST), Andrea Arcangeli
<andrea@e-mind.com> said:
> Agreed. I just said Chuck that I would have extracted the rbtree patch
> against 2.2.6. I _completly_ agree and right now I am also _very_ worried
> by performances of RB in the buffer cache. The rb-cache per-inode is
> almost sure a win according to me,
I don't see that as such a sure thing.
With a page cache hash, we can dynamically increase the hash table size
with memory size to keep a bound on the expected number of pages per
hash bucket. That, plus the natural low complexity of hashing, is good
for predictable performance.
Once you have something like Oracle or an LDAP server serving out
multi-megabyte data files, the equivalent btree is going to get awefully
deep awefully quickly, and there's no simple way to expand the width of
the tree dynamically to minimise that cost.
In other words, I can throw memory at hashing to make it faster, but
trees have a fixed cost which necessarily grows with their size.
--Stephen
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/