Hi
We have a forum site that is behind a load balancer and the site relies
on ip addresses to detect failed logins. This does not work because the
loadbalancer changes the source ip address to that of itself.
our load balancer(f5) adds the original ip as x-forwarded-for http
header.
However the t
> -Original Message-
> From: André Warnier [mailto:a...@ice-sa.com]
> Sent: 02 January 2009 10:00
> To: users@httpd.apache.org
> Subject: Re: [us...@httpd] Isolating slow and fast connections using
> apache/modjk
>
> gerhardus.geldenh...@gta-travel.com wrote:
> [...]
> Not a real answer to
Hi
We have an apache server that load balances two types of applications
across different stacks of tomcats.
We have a tomcat stack for requests that processes very quickly, lets
call this stack A and a tomcat stack for slower running
request(different type of application) named stack B.
We limi
Hi Lesly
I don't know if you have ever watched how they balance a car wheel, but
if not. They bolt it to an electrical motor and spin it around. There is
sensors on this motor that detect where weight needs to be added to get
the wheel balanced so it wont vibrate when driving above 100km/h.
What I
Hi Leslie,
Have you read the documentation on these settings?, it is fairly
helpfull in getting a basic idea of what to tune.
Other than that it really depends on what your apache server is serving.
Is it static pages, does it connect to tomcat, what type of connectors
does it use, php, cgi etc.
Hmmm,
thanks that did the trick.
I am not sure what the differences is between ibm httpd and centos/redhat httpd
but I dont get the error with the ibm version and it still logs succesfully. I
will have to do some more testing to get to the bottom of the problem.
Thanks again, regards
> -
Thanks!
I have changed it to:
CustomLog "|/usr/bin/logger -p local5.warn" combined
which has solved my problem on my test system.
How would you do logger -t ''?
There is a subtle difference in using an empty tag and not using a tag
at all. I will have to check but I suspect the upstream log an
Hi André
Doing cat error.log | /usr/bin/logger -t '' -p local5.warn works perfectly
fine. I did'nt use error log as it is too big and just created a temp file with
unique entries to prevent syslog from saying previous message repeated x times.
logger I believe is part of syslog. In centos's cas
Hi
I am trying to get a simple piped logging example working
Entry in httpd.conf:
CustomLog "|/usr/bin/logger -t '' -p local5.warn" combined
Apache version:
Server Version: Apache/2.2.3 (CentOS)
Server Built: Jan 15 2008 20:36:04
Worker MPM
I am getting the following errors in the error log:
pip
Hi Dimitry,
That sounds more like a firewall related question than apache. You
probably want to configure natting on your firewall or just open up port
80 on the firewall, and then making sure that your apache configuration
security wise is up to scratch.
Regards
> -Original Message-
> F
Hi Johan,
>
> You can check out my whole conf here :
> http://www.pastebin.ca/704464 I'm still wondering how to
> setup ServerLimit / MaxClients bigger then 256 *IF* this
> option are the best ones to increase the typical answer time
> of my server ?
You need to set serverlimit in you confi
Hi Johan,
In you original post ServerLimit was not part of your config, could you
just to make sure paste your current config on the list again.
Also make sure you are using prefork and not mpm otherwise different
rules will apply in configuring MaxClients and ServerLimit.
On a redhat system you
Hi Johan,
> FINALLY, my question is :
>I would like to allow more machine resources to Apache to
> handle more connections and reduce the time to display a page.
>I reduce the timeout to 45 and :
>
> StartServers 15
> MinSpareServers
Hi James
Try adding the following line to your mod_jk config for more verbose
logging:
JkRequestLogFormat "%w %V %U %s %T %B %H %m"
It can be a few things that cause the errors you are getting. I would
suggest reading the timout howto on mod_jk to set a few settings:
http://tomcat.apache.org/co
The reply is a bit late for your problem now, but another good way is to
to use strace to attach to the apache process that is consuming the most
resources at a particular time. It will give you absolutely loads of
information.
Regards
> -Original Message-
> From: Samuel Vogel [mailto:[EM
Hi
Does anyone know if there is an easy way to ensure that a certain
webpage can always be served to a specific IP address even if
maxclientconnections have been reached.
We have an apache server with a low amount of connections. We use
server-status and jkserver-status to monitor the state of th
Hi
Platform: Linux LONGAPA02C 2.6.9-55.ELsmp #1 SMP Wed May 2 14:04:42 EDT
2007 x86_64 x86_64 x86_64 GNU/Linux
Httpd: httpd-2.0.52-32.3.ent.centos4
mod_jk .25
When using mod_jk it complains about this file not being present
No JkShmFile defined in httpd.conf. Using default
/etc/httpd/logs/jk-r
Hi
Not sure if it is related but we have seen problems before that when you
restart apache very often under load it runs out of semaphore. That
might worth looking into.
To clean up semaphores after stopping apache you can try the following:
ipcs -s | grep www-data | awk '{print $2}'|xargs -n 1 ip
The benefit in quick drives is relative. It would depend on your
application.
If you are serving huge files all of the time then yes, and probably
also if you are saving a lot of files most of the time.
Linux caches disk reads so having lots of ram would help with this, but
I do not know which OS
With that type of load that server would be ample. Maybe buy an
additional network card and configure fail over between the cards...
Something to consider is what is your expected growth over the lifetime
of this server. If you depreciate the servers cost to 0 in 3 years time
and expect to replace
Hi
We are using prefork apache on Red Hat Enterprise Linux AS 4 with the
following settings:
MaxKeepAliveRequests 100
KeepAliveTimeout 15
StartServers 8
MinSpareServers8
MaxSpareServers 300
ServerLimit 575
MaxClients 575
MaxRequestsPerChild 4000
What I am unclear abou
21 matches
Mail list logo
