https://bz.apache.org/bugzilla/show_bug.cgi?id=63626

--- Comment #8 from Michael Osipov <[email protected]> ---
(In reply to Yann Ylavic from comment #6)
> AIUI, the purpose of the 100-continue mechanism is to ping the service, as
> opposed to ping the (TCP-)connection.
> In this case the client sent the HTTP header with "Expect: 100-continue",
> mod_proxy_http forwarded it (TCP connect/reuse works) but did *not* respond
> with the interim "100 continue" message yet (since httpd-2.4.41 and bug
> 60330 ;) ).
> So no body has been engaged yet and the service can be considered
> unavailable, i.e. the client or mod_proxy can send the non-idempotent
> request elsewhere since the previous attempt was consciously incomplete.

How am I then supposed to differ between gateway timeout and service
unavailable exactly in the case you have described for bug 60330? I understand
that the ping is different (internal) in nature, but technically the same. As a
server admin I must be able to see a difference in the logs for this.

-- 
You are receiving this mail because:
You are the assignee for the bug.
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to