Auto-loading Kernel Modules

Removing code from the kernel that provides unneeded support is one option long associated with Linux. Now, you can remove code that is not constantly required, putting it in modules loaded on command.
What Should Be Compiled as a Module?

When configuring your system, at first you may think it would be best to compile everything as a module. This isn't always a good idea, as it won't always save memory. Each module uses memory in 4K pages, so the last page will generally have some space wasted. Therefore, if you'll almost always have the module in use, you might as well compile it into the kernel. Also, keep in mind that kerneld itself consumes some memory (in my experience, at least 12 pages), so if you only have a few small modules to worry about, it would be better to compile them into the kernel or load them explicitly in the startup script.

Modules for file systems must be loaded as long as the file system is mounted, even if you're not using it. So if you keep /dos mounted all the time, don't bother to compile support for FAT as a module. If you don't like that option, you could look into using an automount daemon instead of keeping the file system mounted.

Be careful with modules that include information that may be changed when run. For example, the sound driver keeps track of the volume, and if you compile it as a module, the volume will be reset to the default each time it is loaded.

Finally, be careful not to compile something as a module if it will be used at boot time before kerneld is started. This includes the root file system, of course. For many systems, you'll find that you need both ELF and a.out support before kerneld starts. You may be able to overcome some problems by installing kerneld as one of the first programs executed by the startup scripts, but be careful if you're also doing dial-on-demand, as you may have something like sendmail in your startup scripts that will trigger it. As long as you have your old kernel around as a safety net, though, feel free to experiment.

Preston Crow Preston Crow is a graduate student in computer science at Dartmouth College. He became a happy Linux user in the summer of 1995, shortly before becoming happily married.

______________________

Webcast
How to Build an Optimal Hadoop Cluster to Store and Maintain Unlimited Amounts of Data Using Microservers

Realizing the promise of Apache® Hadoop® requires the effective deployment of compute, memory, storage and networking to achieve optimal results. With its flexibility and multitude of options, it is easy to over or under provision the server infrastructure, resulting in poor performance and high TCO. Join us for an in depth, technical discussion with industry experts from leading Hadoop and server companies who will provide insights into the key considerations for designing and deploying an optimal Hadoop cluster.

Learn More

Sponsored by AMD

White Paper
Red Hat White Paper: Using an Open Source Framework to Catch the Bad Guy

Built-in forensics, incident response, and security with Red Hat Enterprise Linux 6

Every security policy provides guidance and requirements for ensuring adequate protection of information and data, as well as high-level technical and administrative security requirements for a system in a given environment. Traditionally, providing security for a system focuses on the confidentiality of the information on it. However, protecting the data integrity and system and data availability is just as important. For example, when processing United States intelligence information, there are three attributes that require protection: confidentiality, integrity, and availability.

Learn more about catching the bad guy in this free white paper.

Learn More

Sponsored by DLT Solutions