Re: HTCP versus ICP

On Tue, 30 Jun 1998, Dax Kelson wrote:

> http://www.vix.com/ietf/htcp.txt
> Is HTCP a "good" thing?

HTCP is essentially ICP except that HTTP headers are included with ICP_HIT
reply. Knowledge of HTTP headers may help to reduce false hit ratio (peer
will not request objects that appear to be "stale" from peer's point of
view).

No protocol can guarantee no false hits though. Personally, I think that a
small false hit ratio is acceptable as long as Squid can handle false hits by
requesting the missing object from a parent or going direct. The latter is
being implemented in Squid now.

HTCP does not solve scalability problems of ICP, as far as I can see. A
cache has to wait for miss replies from all peers before going through a
parent or direct.

> Will it be implemented in Squid 1.2?

Initial implementation of HTCP got stuck mainly for performance reasons.
HTCP, among other things, would require disk I/Os to swap in object headers
for every ICP_HIT reply because Squid does not keep all headers in memory.
This does not answer your question though.

Regardless of HTCP status, _if_ most of us consider small false hit ratio
acceptable, Cache Digests is the way to go. They do not introduce any
query/response delays and, thus, scale well with the number of peers. The
price is more RAM, but with less than a 1MB digest per 16GB peer and cheap
memory, it's not a big deal.

$0.02,

Alex.
Received on Tue Jun 30 1998 - 22:27:11 MDT