httpd-bugs mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From bugzi...@apache.org
Subject DO NOT REPLY [Bug 37770] - proxy: error reading status line from remote server (null)
Date Thu, 06 Dec 2007 21:21:59 GMT
DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG·
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
<http://issues.apache.org/bugzilla/show_bug.cgi?id=37770>.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND·
INSERTED IN THE BUG DATABASE.

http://issues.apache.org/bugzilla/show_bug.cgi?id=37770





------- Additional Comments From rpluem@apache.org  2007-12-06 13:21 -------
(In reply to comment #67)
> (In reply to comment #65)
> > All these messages appearing in error.log indicate some problem with network on

> > server side or with server itself and do not mean errors in proxy software. I 
> > think the bug should be closed.
> 
> What data are you privy to that causes you to believe that there isn't a bug
> causing the proxy module to believe one of those conditions is true when it, in
> fact, is false?

He is correct in several cases:

(70007)The timeout specified has expired -- indicates network timeout.

This can happen if your backend takes a long time preparing the response and
this is a wanted behaviour. You can adjust the timeout either via TimeOut,
ProxyTimeout or a specific worker timeout.

(104)Connection reset by peer -- TCP RST was sent from server
(70014)End of file found -- TCP FIN was sent from server

These can happen either because of network problems or problems on the backend
server or because of a race condition in the proxy code.
Currently there is a race condition that the backend server can close the
keepalive connection right after httpd checked the state of this TCP connection
and httpd starts sending a request and waiting on a response on a connection
that is in fact dead. But this should happen rarely and browsers try to resent
the request in such situations provided the request was idempotent (so no POST
request or GET request with parameters).
All other cases (except for the race condition) also lead to error messages in
the browser (maybe a different one that the one sent by the proxy, but an error
message).
So the race condition is the only reason why I think that it is justified to
keep this PR open.



-- 
Configure bugmail: http://issues.apache.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

---------------------------------------------------------------------
To unsubscribe, e-mail: bugs-unsubscribe@httpd.apache.org
For additional commands, e-mail: bugs-help@httpd.apache.org


Mime
View raw message