Load balancing with the Apache http server
Balancing Act

© sTan, Fotolia
Today's web performance and availability requirements make load balancing indispensable. In this article, we show you how to set up an effective load balancing system using features built into the Apache web server.
A multitude of technologies support load balancing for web servers. Load balancers come in all shapes and sizes, from simple DNS-based techniques through vast and versatile proprietary systems. In some cases, however, the load balancing features you need might be available already through the Apache web server. In this article, I describe some strategies for load balancing with Apache.
Preparation
The schematic in Figure 1 shows the underlying structure of a software-based balancing system. In this scenario, several front-end load balancers accept incoming user requests and distribute them to a pool of back-end servers on the basis of a predefined scheme. Multiple individual systems can run in parallel to provide a fail-safe (shown in the background in Figure 1). Apache includes a number of modules for supporting load balancing (Table 1), and you'll need to make sure any modules you intend to use are loaded:
LoadModule xyz_module modules/mod_xyz.so
As you can see in Table 1, Apache's basic load balancing capabilities include features such as caching, compression, URL rewriting, and header processing. Some of the modules in Table 1 are loaded by default.
Consult your own Apache configuration for more on which modules might already be present on your system.
If JServ-capable application servers, such as Apache Tomcat or Jetty, are used, the gateway can also use the Apache JServ Protocol (ajp). All you need to do is load the mod_proxy_ajp module instead of mod_proxy_http and change the URLs from http:// to ajp://.
The use of this binary-format protocol offers a couple of advantages with regard to back-end connection performance and lower resource overheads, but this functionality is bought at the price of more permanent connections to the back ends.
Incidentally, you can run ajp and http back ends at the same time as members of the same pool. For the sake of completeness, keep in mind also that Apache supports ftp proxy with the mod_proxy_ftp module.
For additional details, check out the Apache http server documentation [2].
Configuration
The sample configuration shown in Listing 1 includes the basic front-end server settings for load balancing in Apache.
Listing 1
Sample Configuration
This configuration starts with ProxyRequests to disable the normal proxy mode and setting up what is known as a reverse proxy, or gateway, to be more precise. Disabling Via headers (ProxyVia) makes the gateway invisible.
The ProxyPreserveHost and ProxyErrorOverride commands ensure that the host headers included in the request are passed on to the back ends and that any error messages generated by the back ends are replaced by the load balancer and thus standardized. The output of a suitable timeout, with ProxyTimeout, rounds off the basic configuration.
The core definition, that of a back-end pool and its members, is handled by a Proxy container and the specification of a special balancer:// schema followed by the pool name. The BalancerMember instructions and parameters in the container specify the individual members along with their properties.
At the end of the configuration, the back-end pool defined previously is assigned a separate URL space; more parameters define the load balancer's generic approach. To enable regular expressions, you could use the advanced ProxyPassMatch command instead of ProxyPass.
As an alternative, the rewrite module (mod_rewrite) and custom rules would unleash the full power of regular expressions. However, in this case, you will need to use ProxySet because load balancer parameters cannot be modified by rules:
ProxySet balancer://pool1 lbmethod=bytraffic ... RewriteEngine On RewriteRule ^/+(.*)$balancer://pool1/$1 [P,L]
Listing 1 thus defines two back-end servers for pool1. Requests are distributed on the basis of the number of requests (see the lbmethod parameter). The load factor setting assigns twice as many requests to server1 compared with server2. Connections are reused but also restricted to a maximum value. The URL space is defined as the complete URL space below /shop.
Table 2 provides a summary of the most common ProxyPass and BalancerMember commands. For more information, see the Apache http server documentation [2].
Options
The Apache http server's proxy module (mod_proxy) provides an unbelievable range of special settings. Tools are available for many different scenarios. For example, you can use the status parameter to operate a hot standby server:
BalancerMember http://server4:1080... status=+H
This command specifies that server4 is only enabled if all the remaining pool members fail.
This server is the last line of defense and can be used to serve up a restricted version of a web application or to forward requests to a substitute system.
Buy this article as PDF
(incl. VAT)
Buy Linux Magazine
Subscribe to our Linux Newsletters
Find Linux and Open Source Jobs
Subscribe to our ADMIN Newsletters
News
-
KDE Plasma 6 Looks to Bring Basic HDR Support
The KWin piece of KDE Plasma now has HDR support and color management geared for the 6.0 release.
-
Bodhi Linux 7.0 Beta Ready for Testing
The latest iteration of the Bohdi Linux distribution is now available for those who want to experience what's in store and for testing purposes.
-
Changes Coming to Ubuntu PPA Usage
The way you manage Personal Package Archives will be changing with the release of Ubuntu 23.10.
-
AlmaLinux 9.2 Now Available for Download
AlmaLinux has been released and provides a free alternative to upstream Red Hat Enterprise Linux.
-
An Immutable Version of Fedora Is Under Consideration
For anyone who's a fan of using immutable versions of Linux, the Fedora team is currently considering adding a new spin called Fedora Onyx.
-
New Release of Br OS Includes ChatGPT Integration
Br OS 23.04 is now available and is geared specifically toward web content creation.
-
Command-Line Only Peropesis 2.1 Available Now
The latest iteration of Peropesis has been released with plenty of updates and introduces new software development tools.
-
TUXEDO Computers Announces InfinityBook Pro 14
With the new generation of their popular InfinityBook Pro 14, TUXEDO upgrades its ultra-mobile, powerful business laptop with some impressive specs.
-
Linux Kernel 6.3 Release Includes Interesting Features
Although it's not a Long Term Release candidate, Linux 6.3 includes features that will benefit end users.
-
Arch-Based blendOS Features Cool Trick
If you're looking for a Linux distribution that blends Linux, Android, and web apps together, blendOS might be what you're looking for.