Re: Patch to make Reload into IMS work more sanely

From: Henrik Nordstrom <hno@dont-contact.us>
Date: Sat, 15 Aug 1998 13:43:53 +0200

Duane Wessels wrote:

> How about if we change
>
> TCP_IMS_HIT --> TCP_IMS_NM
> TCP_IMS_MISS --> TCP_IMS_HIT

I have said this earlier, but here it is again:

Why have two log tags? The IMS outcome is clearly visible as the HTTP
code.

TCP_IMS_HIT is always 304
TCP_IMS_MISS is never 304

TCP_MISS is sometimes 304 (depending on client request and server
status)

I think TCP_IMS_* should be replaced by the single tag TCP_HIT_IMS.

Why:
1. It is a HIT in all cases.
2. The IMS status is available as the HTTP code, in the same way as on
TCP_MISS, thus only one log tag is needed and is less confusing.
3. The only reason why to mention IMS at all is to make it possible to
build vague statistics on the efficiency of downstream caches (proxies
and client caches).

/Henrik
Received on Tue Jul 29 2003 - 13:15:51 MDT

This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:11:51 MST