Hello,
After cleared the browser's cache, we observed multiple user-tracking
cookies generated when the request proxies to another env. The capture header
shows,
In Response Header
CP_GUTC:10.37.138.15.1400173959753313
In Request Header
CP_GUTC:10.37.145.95.1400173959439209
we
Team,
After cleared the browser's cache, we observed multiple user-tracking
cookies generated when the request proxies to another env. The capture header
shows,
In Response Header
CP_GUTC:10.37.138.15.1400173959753313
In Request Header
CP_GUTC:10.37.145.95.1400173959439209
we
hello experts,
We see multiple User-Tracking cookies created in a request's header, when it
was routed to cross multiple domains. Is there a way we can configure the
Apache to not create another user-tracking cookie if it is already created?
Many Thanks,
Q.Xie
Hello Team,
Our proxy load-balances two members showed as below. How to configure the
probe to probe each of them and failover if one member is down? I think I may
need to configure ProxySet nofailover=Off explicitly. Is nofailover=Off the
default? Also, is there any other or better
Hello,
I'd like to filter some non-business requests into a separate log file. Is
there any a way to do it?
Thanks, Q.Xie
-
To unsubscribe, e-mail: users-unsubscr...@httpd.apache.org
For additional commands, e-mail:
Hello,
according to Apache
website http://httpd.apache.org/docs/2.0/mod/core.html#timeout, it says,
The TimeOut directive currently defines the amount of time Apache will wait for
three things:
1. The total amount of time it takes to receive a GET request.
2. The amount of
Hi,
Our env has Apache-2.2.19 + Openssl-1.0.0d compiled in a Linux host, it is
configured as a ssl proxy server. When Client-Cert-Authentication enabled, it
triggered Segmentation Fault. This happens on Openssl-1.0.0 other
versions(a,c). However, Client-Cert-Auth is working fine with
Hello,
I'd like to compile a 32-bit Apache 2.2.17 on a 64-bit Red Hat Linux host.
Could someone tell me how to do it?
Many Thanks,
Q.Xie
-
The official User-To-User support forum of the Apache HTTP Server
Hello,
we have Apache 2.2.17 + Openssl-1.0.0a compiled under a RedHat Enterprise
Linux 64-bit host. It's installed in a Reverse Proxy Server, which ssl proxy
to
backend. It failed if configured with Client-Cert-Auth. Turning on log level
to Debug, it threw errors pasted below. I
Hello!
We have a Reverse Proxy server, which initiates SSL connection to the backend
server, functioning as a SSL client. In order to enable Session-Resumption,
should we configure SSLSessionCache and SSLSessionCacheTimeout?
Thanks,
Q.Xie
Hello,
In a Reverse Proxy server, what is the relation of Apache TimeOut and
ProxyTimeout? Can someone shed some lights on it?
Thx, Q.Xie
-
The official User-To-User support forum of the Apache HTTP Server
the observation of my server, the whole HTTP session won't be terminated
till 10 min. I am not sure why since the front connection should be
terminated in 5 min?
Thx, Q.Xie
- Original Message
From: Torsten Foertsch torsten.foert...@gmx.net
To: users@httpd.apache.org
Cc: Qingshan Xie xieq
- Original Message
From: André Warnier a...@ice-sa.com
To: users@httpd.apache.org
Sent: Wednesday, August 19, 2009 2:00:26 PM
Subject: Re: [us...@httpd] ProxyTimeout
Qingshan Xie wrote:
Thanks for your quick reply.
I am wondering what will happen in the case below,
Timeout
Hello!
Our Reverse Proxy servers proxy everything to backend generally. How can we
configure it not proxy some specific URIs or Files, for example, the local
index.html?
Thanks,
Q.Xie
-
The official
Hello,
I am pretty confused about the TCP connections and sessions in the case of
proxy server in the middle. Here is the case,
Browser - SSL-Proxy Server - SSL WebServer
there are two parts in this SSL communication,
1. Browser - SSL-Proxy Server. Browser sends a
Hello!
I tried to set a new variable in the request header, for example $A, which
copies another header variable $B's value. I tried to use mod_headers and
mod_evn to set it but none of the success. Do you have any good ideas? Please
let me know.
Appreciation!
Q.Xie
I tried to use SetEnv to set the REMOTE_ADDR but failed.
Our Apache webServer has a network device in front, which NATed the client
sourceIP. In order to retrieve the sourceIP, we used network device inserted
the sourceIP, like to use the inserted one to replace the REMOTE_ADDR. I tried
to
Team,
As you know filer or shared storage has been widely used in the
infrastructure build. In the case of multiple web hosts share the same filer,
does Apache instance on a host need to read/write some running information into
the local drive, which are not sharable on filer?
Thanks,
NULL-SHA
You
also
need
to
make
sure
that
your
backend
server
is
configured
to
accept
NULL
ciphers.
-ascs
-Message
d'origine-
De
:
Qingshan
Xie
[mailto:[EMAIL PROTECTED]
Envoyé
:
jeudi
31
janvier
2008
01:37
À
:
Krist
van
Besien;
users@httpd.apache.org
Objet
From: Krist van Besien [EMAIL PROTECTED]
To: users@httpd.apache.org
Sent: Friday, January 25, 2008 1:18:54 AM
Subject: Re: [EMAIL PROTECTED] Can Apache Proxy server to proxy the HTTP
requests to the backend HTTPS/SSL server?
On
Jan
25,
2008
12:22
AM,
Qingshan
Xie
[EMAIL PROTECTED]
wrote
, 2008 12:22 AM, Qingshan Xie [EMAIL PROTECTED] wrote:
Dear Friends,
We configured a HTTPS proxy server successfully to server the
HTTPS requests. However, we also want to configure a HTTP proxy server to
handle the HTTP requests but proxy the HTTP requests to the backend
SSL(or HTTPS
This maybe a little crazy or sounds silly. We have a reverse proxy talks to
the backend Apache in SSL. However for better performance we wonder if we can
disable the request encryption and decryption but only ssl-handshake? This
pretty much like SSH pipe, once the trust-relation
Dear Friends,
We configured a HTTPS proxy server successfully to server the HTTPS
requests. However, we also want to configure a HTTP proxy server to handle the
HTTP requests but proxy the HTTP requests to the backend SSL(or HTTPS) server.
The request flow is as below,
HTTP request ==
I could compile mod_security-1.9.4 with Apache-2.0.61 on Solaris 8 easily but
failed on 2.1.4. Seems 2.1.4 compilation is very different from 1.9.4. If
some one has such experience, could you share it with me?
Many Thx,
Q.Xie
Hello,
There is a need that an Apache server converts a HTTP request to HTTPS, and
proxies it to the backend HTTPS server. Can an Apache instance do it?
Many Thanks,
Q.Xie
Get easy, one-click
Hello! Apache expert,
If a Reverse Proxy (RP) talks to another Apache
server(AS), is there a way to do bidirectional
authentication or Client-Cert-Auth between those two
Apache servers? It means, the RP needs to send the
client cert to AS for ssl-handshake. Can we do it?
Many Thanks,
hello,
we have a 2.0.59 running on Solaris 8. We
configured mod_ssl SSLSessionCache and SSLMutex as
below,
SSLSessionCache
dbm:/opt/httpd/logs/httpd-ssl-443/ssl_scache
SSLSessionCacheTimeout 300
SSLMutex file:/opt/httpd/logs/httpd-ssl-443/ssl_mutex
However, we never see the log
Thanks William for your quick reply.
I set the path for ssl_mutex log in local and correct
permission mode, however, I still could not see the
ssl_mutex log created after restart and tests. any
idea?
Thx, Q.Xie
--- William A. Rowe, Jr. [EMAIL PROTECTED]
wrote:
Qingshan Xie wrote:
hello
Hi,
After Apache 2.0.59 KeepAlive On, IE6 users got
Page cannot be found or Page cannot be displayed
errors, but Firefox users worked fine. Do you know
what is the root cause?
Many Thx,
Q.Xie
Be
I have two questions on Sessions and Connections.
1. If a browser sends multiple requests and
establishes Two Connections with an Apache server, how
many sessions it will be generated, one or two?
2. If a browser sends multiple HTTPS requests to two
load-balanced Apache servers, how many
I still see the default configuration for Apache
2.0.59 HTTPS/SSL server setup,
SetEnvIf User-Agent .*MSIE.* \
nokeepalive ssl-unclean-shutdown \
downgrade-1.0 force-response-1.0
Does MSIE still have the issue to use Apache KeepAlive
feature in HTTPS, even MSIE 6 and 7? Is
Hi,
After KeepAlive On, multiple requests will be handled
one at a time, or say Sequentially. If there are 30
pages requested, could this Sequential
Processing(like a queuing process) slow down the
performance? On the other hand, without KeepAlive, 30
requests could be handled by 30
Thanks Joshua for your quick reply and help!
One more quick question on KeepAlive,
Can KeepAlive work on POST?
Thx, Q.Xie
--- Joshua Slive [EMAIL PROTECTED] wrote:
On 6/6/07, Qingshan Xie [EMAIL PROTECTED] wrote:
Hi,
After KeepAlive On, multiple requests will be
handled
one
Hello All,
We have apache 2.0.59 on Solaris 8. It sometimes
took ~120 seconds to return a 404. I don't understand
why took so long. Can someone help me to understand
it? If convenient, could you also tell me how apache
return the 404?
Many Thx
Q.Xie
Hi! all,
In Apache/2.0.59, if I need to specify multiple
paths for ProxyPass and ProxyPassReverse, for
example, I need to proxy 3 paths /A, /B, /C to
http://www.foo.com
do I need to configure them as,
..
ProxyPass: /A http://www.foo.com
ProxyPassReverse: /A http://www.foo.com
ProxyPass:
Thanks a lot for your quick reply!
Yes, we are setting a reverse proxy. Have not found a
simple way to define the multiple paths.
Q.Xie
--- Dragon [EMAIL PROTECTED] wrote:
Qingshan Xie wrote:
Hi! all,
In Apache/2.0.59, if I need to specify multiple
paths for ProxyPass
Hi,
I am not sure how good Apache mod_proxy on its
Load-Balance and Failover, in regarding its
Performance, Capacity Limit, Security, etc? Can
someone help me?
Thx, Q.Xie
Yahoo! Music Unlimited
Access
Hi! All,
I'd like to configure a https proxy server to route
the request from https://A.com/xyz/ to
https://B.com/xyz/. Can Apache mod_proxy do it and
how to configure it?
Many Thanks in Advance,
Q.Xie
Hi,
After compiled Apache-1.3.37 on a Solaris 2.6
host,it's fine to run httpd -l. However, after
installed it on another solaris 2.6 host, run httpd
-l prompted error httpd: Unable to exec Shared Core
Executable Program
`/opt/httpd/apache/1.3.37/lib/libhttpd.ep'. Can some
tell me why?
Many
Hi!
In worker MPM, the default ThreadsPerChild is 25.
As I understand, there should be one thread is the
Listener. So, the real number of connections/requests
a child can handle should be 24 not 25. Am I right?
Thx, Q.Xie
__
Do You Yahoo!?
Hi,
I read the docs at
http://httpd.apache.org/docs/1.3/keepalive.html which
indicates that,
However, Keep-Alive support only is active with files
where the length is known beforehand. This means that
most CGI scripts, server-side included files and
directory listings will not use the
Hi,
I read the docs at
http://httpd.apache.org/docs/1.3/keepalive.html which
indicates that,
However, Keep-Alive support only is active with files
where the length is known beforehand. This means that
most CGI scripts, server-side included files and
directory listings will not use the
Hi All,
I am not clear about the role of the Apache parent.
I know the parent do control the children; but I am
not clear if the parent also do the service as below,
Listen to the port(80)
Accept the connection
Establish the Connection
Dispatch the Connection to a idle child
If
--- Joshua Slive [EMAIL PROTECTED] wrote:
.
No. The children are responsible for grabbing
connections. See the
documentation on AcceptMutex and the performance
tuning docs for a detailed discussion of how this
happens.
Thanks Joshua!
Where those IDLE children to get the AcceptMutex
Hi! All,
I did a count. there are 192 moduels in 2.0.59
source. are all those modules production modules?
the production modules mean those modules have been
thoroughly tested without issues. Do we have a list
of modules most widely used?
Thanks, Q.Xie
Hi! All,
I think MPM(prefork/worker)talks to OS for
process/threads management, but I am not clear if
Apache Core also talks to OS for network
connections(sockets) etc? Can some help me to clarify
the correlation between Apache Core and MPM?
Thx, Q.Xie
My question may be silly but I'd like to know it. If
the number of Idle Threads is over the
MaxSpareThreads, how Apache parent clean those extra
threads? clean the threads or kill the children?
Thx, Q.Xie
__
Do You Yahoo!?
Tired of spam? Yahoo!
Thanks Joshua! I did mean worker.
Q.Xie
--- Joshua Slive [EMAIL PROTECTED] wrote:
On 10/18/06, Qingshan Xie [EMAIL PROTECTED] wrote:
My question may be silly but I'd like to know it.
If
the number of Idle Threads is over the
MaxSpareThreads, how Apache parent clean those
extra
in this case?
Thx, Q.Xie
--- Joshua Slive [EMAIL PROTECTED] wrote:
On 10/18/06, Qingshan Xie [EMAIL PROTECTED] wrote:
My question may be silly but I'd like to know it.
If
the number of Idle Threads is over the
MaxSpareThreads, how Apache parent clean those
extra
threads? clean
] wrote:
On 10/18/06, Qingshan Xie [EMAIL PROTECTED] wrote:
Yes. It's worker mpm. My concern is if the
worker
configured as below,
MinSpareThreads 10
MaxSpareThreads 20
ThreadsPerChild 50
and in the case there are only 10 connections to
that
child, but 40 threads are idle
# nokeepalive ssl-unclean-shutdown \
# downgrade-1.0 force-response-1.0
but I am not sure what is the negative impact of this
disabling. If you know it, please let me know.
Many Thanks, Q.Xie
--- Joshua Slive [EMAIL PROTECTED] wrote:
On 10/14/06, Qingshan Xie [EMAIL PROTECTED] wrote
--- Joshua Slive [EMAIL PROTECTED] wrote:
On 10/13/06, Qingshan Xie [EMAIL PROTECTED] wrote:
Hi! All,
We have a 2.0.59 HTTPS server on Solaris 8.
After
I turned on KeepAlive On and monitored it by
mod_status, I did not see any connection in
KeepAlive
state. Seems keepalive
PROTECTED] wrote:
HTTP is sessionless protocol so it needs KeepAlive
feature to keep TCP
connection active. HTTPS connections are always
active untill client
leaves your site.
On 10/13/06, Qingshan Xie [EMAIL PROTECTED] wrote:
Hi! All,
We have a 2.0.59 HTTPS server on Solaris 8.
After
Hi! All,
We have a 2.0.59 HTTPS server on Solaris 8. After
I turned on KeepAlive On and monitored it by
mod_status, I did not see any connection in KeepAlive
state. Seems keepalive is not applicable on HTTPS/SSL
server. Can some one tell me if this is true and why?
Thx, Q.Xie
If my apache KeepAlive On, multiple hits will go to
the server via the same connection. In this case, how
apache count the hits if I keep hit URLs to the same
server?
I did a test when the KeepAlive turned on, the child
did not exit after I clicked the urls over 20 times
while my
Thanks Joshua. that url answered my question.
I have another question about MaxSpareThreads. If
the number of spare threads is over MaxSpareThreads,
will Apache parent kill the spare child? If the
number of hits to that spare child is still below the
limit of MaxRequestsPerChild, will Apache
Hi,
We have Apache-2.0.54 with worker MPM in Solaris 8,
the confgiuration part about worker.c is as below,
IfModule worker.c
ServerLimit 3
ThreadLimit 250
StartServers 1
MaxClients 750
MinSpareThreads 100
MaxSpareThreads 350
ThreadsPerChild
Thanks Joshua for your help. Please see my
reply/comment inserted below.
Thx again, Q.Xie
How are you measuring? ServerLimit is a limit on
the number of processes, not threads. On many OSes,
it is
difficult to distinguish between processes and
threads.
Our WebServer is a SUN box with
Hi! Joshua, Under the high load condition, if the
number of children spawned is over the ServerLimit or
MaxClients/ThreadsPerChild, what the status of those
extra children is? Will all those extra children be
in the queue waiting for the process?
Thx, Q.Xie
--- Joshua Slive [EMAIL PROTECTED]
Apache will never create children over
ServerLimit/MaxClients.
Joshua, I did see our apache created extra children
spawned by Apache parent. We configured Apache-2.0.54
with worker MPM, the serverlimit was 32 but it spawned
over 42 children in the high load condition. Any
comment?
Thx,
Joshua,
Sorry for late question.
I thought mod_dir will append this trailing-slash but
seems Apache core can also do it. what is the
relation of mod_dir and Apache core for this
trailing-slash redirect handling?
Thx, Q.Xie
--- Joshua Slive [EMAIL PROTECTED] wrote:
..
This is called a
Hi! All,
We have Apache-2.0.54 compiled with worker mpm.
After load mod_ldap and mod_auth_ldap as below in
httpd.conf,
LoadModule ldap_module modules/mod_ldap.so
LoadModule auth_ldap_module modules/mod_auth_ldap.so
LDAPSharedCacheSize 0
LDAPCacheEntries 1024
LDAPCacheTTL 600
Hi! All,
We have an Apache-2.0.54 instance running on a SUN
box of Solaris 8, it runs as worker MPM. I tried to
increase its ThreadsPerChild for our env requirement.
Seems it's OK to push the limit to 350 ThreadsPerChild
but failed it's higher than 400. The processes died
automatically
--- Joshua Slive [EMAIL PROTECTED] wrote:
..
Having a portion of children hanging out in
keepalive
could hurt even if there are children free, because
it could cause more context switches/memory cache
misses when switching requests between children.
I don't fully understand what you
Hi,
I am curious how Apache manages all children, which
of them are in the service to clients, which of them
are Idle, etc.? Can some one give me a brief
description or refer me a URL/books?
Many Thanks, Q.Xie
__
Do You Yahoo!?
Tired of
Joshua, I re-read your comment, have couple questions
inserted below. Could you please help me again?
Thx, Q.Xie
--- Joshua Slive [EMAIL PROTECTED] wrote:
..
If the load-testing clients are not using
keep-alives, the server will
almost certainly be slower with keep-alives turned
on,
multiple
requests by multiple children simultaneously in
parallel. Do you think the above explanation makes
sense?
Thx, Q.Xie
--- Joshua Slive [EMAIL PROTECTED] wrote:
On 8/10/06, Qingshan Xie [EMAIL PROTECTED] wrote:
Hi,All,
We have apache-2.0.54 installed on Solaris and
Linux boxes
Hi,All,
We have apache-2.0.54 installed on Solaris and
Linux boxes. We did a bounch of Load-Test with
KeepAlive On or Off on multiple jpg, gif, text files.
The test results are discourage. The webServer has
slower response time when KeepAlive is On compared to
the condition when KeepAlive
We are using Apache 2.0.54 in a Solaris 8 host or a HP
RedHat, Apache uses 'worker' MPM. Can some one tell me
if there is any issue if the ThreadsPerChild is
increased to 1000?
Many Thanks, Q.Xie
__
Do You Yahoo!?
Tired of spam? Yahoo! Mail has
If the Apache-2.2.2 compiled with mod_so, Apache
should be able to load mod_was_ap20_http.so without
any issue. I think you need to check the permission
bits not only the sub-directory of plugin, but all the
path up to /opt/CHROOT/. You may also need to set the
sticky bit.
Q.Xie
--- [EMAIL
Could some one shed some lights to the email below?
Thanks in Advance. Q.Xie
I also read it, seems the report/patch is pretty
old,applicable to the Apache-2.0.44/2.0.54. Not
sure
if this patch is still suitable for 2.0.58, and
especially if 2.2.2 has already included this
feature?
I also read it, seems the report/patch is pretty
old,applicable to the Apache-2.0.44/2.0.54. Not sure
if this patch is still suitable for 2.0.58, and
especially if 2.2.2 has already included this feature?
Many Thx, Q.Xie
--- Asaf Dalet [EMAIL PROTECTED] wrote:
Hi,
There was a fix for this
Hi, Boyle,
I have a related question. We'd like to implement
a SSL-Login on a HTTP(port 80) webServer to secure the
userId/password. This means, whenever a site needs
the authentication, the webServer redirects it to
HTTPS server for processing. However, this is pretty
annoying since it
and return 304? Do
you know why?
Q.Xie
--- Joshua Slive [EMAIL PROTECTED] wrote:
On 7/5/06, Qingshan Xie [EMAIL PROTECTED] wrote:
Thanks Joshua.
We are using SiteMinder for
Authentication/Authorization, and Single-Sign-On.
The
missing part of the headers is SiteMinder
SessionID
Hi,
Our Linux will be upgraded to 64-bit OS but the
current Apache binary was compiled in 32-bit. Can
32-bit Apache binary run on 64-bit Linux OS without
any issue?
Thx, Q.Xie
__
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam
Hi,
We have an Apache-2.0.54 in front and an WebSphere
in backend. Whenever Apache threw 304 return code. A
lot of info in the header was stripped away. Is there
any a way to keep all those header information when
got 304?
Thx, Q.Xie
__
Do
Hi! All,
We have Apache 2.0.54 installed on a SUN box with
Solaris 8. We use 'worker' MPM. Based on our service
requirement, we'd like to set the
ThreadsPerChild=1000 or even higher. As I know, the
default ThreadsPerChild is 64 in Unix, not sure what's
the impact if the value set is much
William, Thx for your reply and help. -QX
--- William A. Rowe, Jr. [EMAIL PROTECTED]
wrote:
James Benson wrote:
Yes I think I remember seeing something about that
in the apache docs,
the order it ships with should not be changed!
That was true of 1.3.x (the order of all the hooks
All,
I am using Apache 2.0.54 with. The default
SSLCipherSuite is
SSLCipherSuite
ALL:!ADH:RC4+RSA:+HIGH:+MEDIUM:+LOW:+SSLv2:+EXP:+eNULL
In this CipherSuite, I observed nearly all
browsers(IE, Netscape, Firefox) select RC4. Tried to
change it to 3DES, like
SSLCipherSuite
All,
We have a Apache-2.0.52 running on a Linux host.
We use mod_so to load modules appended below. However
I am not very sure if those modules loaded in the
right order or the order matters? Please advise.
Thx, Q.Xie
LoadModule
All,
I am not clear the relation between KeepAlive and
SSL-HandShake. I think if KeepAlive is turned off,
the SSL-HandShake will be processed in every request.
However, if KeepAlive is turned on, the following
requests after the 1st SSL-HandShake during the period
of KeepAlive won't need
Hello,
I have an Apache-2.0.52 installed on a Solaris 8 host.
It is configured to be a worker MPM. Usually Apache
child's RSS is around ~14 MB. However sometimes it
grew to ~230 MB. after run the command
/usr/proc/bin/pmap pid#, it prompted that, the heap
size of the child is around 230 M.
Hi, All,
We have a 2.0.54 Apache with worker mpm running on
Solaris 8. We observed many child processes last for
long time, was not terminated automatically over 8
hrs. This caused too many child processes forked and
running in the system, caused the swap space depletion
problem. wonder if
Hi, All,
We have an Apache-2.0.52 installed in our Solaris
2.8 host. It's a HTTPS server. the configuration is
as below:
==
IfModule worker.c
ServerLimit 22
StartServers 15
MaxClients 500
MinSpareThreads 25
MaxSpareThreads 20
84 matches
Mail list logo
