One of the obvious driving factors around enterprise Linux adoption has been the significant cost savings on software and hardware.

Quite simply, the hardware is cheaper and the OS is cheaper. By taking advantage of the explosion of commodity Linux boxes - and by going the Linux route to get around costly licensing issues on the software side - enterprises are finding the economics around Linux to be quite attractive.

Unfortunately, the cost around managing this "cheaper" IT infrastructure is often overlooked.

Over the last few years, the enterprise has grown increasingly concerned with these costs of IT systems management and administration. In a study put out last year by IDC, for example, the findings confirmed that IT operations management costs are currently growing more than three times faster than the enterprise spend on new servers.

And we're starting to see more discussion around the specific cost/complexity issues of managing Linux systems.

"We are consistently hearing a demand for more mature Linux systems management tools," said Dave Rosenberg, programs director at LinuxWorld Conference and Expo. "There are a lot of companies that are ready to broaden their Linux environments, but are wary about how efficiently they'll be able to manage their new systems."

So what are some of the common challenges that enterprises are facing in managing Linux systems?

Abundance of Servers
In many cases, enterprises are migrating from Unix to Linux - and in the process are replacing large, monolithic boxes (such as the Sun E10K, for example) with scores of commodity hardware, such as Lintel boxes. So the Linux environment typically consists of many more pieces of hardware to manage. Given the dramatic savings in hardware and software costs, commodity computing with Lintel servers should be an obvious choice. Yet, the tidal wave has not yet occurred - while the Lintel hardware and software cost structures are attractive, deploying 10 times as many cheap servers causes an overwhelming administrative burden. The proliferation of Linux servers causes a problem of abundance: the more servers in use, the more differences and interdependencies.

Lack of Sophisticated Tools
There's an overall a lack of maturity on the point of system management and configuration management tools - Linux isn't on a par yet with what's available for Unix. Most Linux systems today are administered through a series of scripts and freeware that are very flexible and give good "hands on" control - but that also take a lot of time to set up and maintain. Scripting and procedural administration that grew up in the 90s Unix boom aren't up to the task of managing hundreds or even thousands of nearly identical servers. The situation gets worse when you want to repurpose servers on-the-fly to accommodate changing workloads.

Version Control and Conflicts
Another factor is the heterogeneity issue - not only do you have a huge number of Linux distributions, but there are varieties within those distributions, and people often tweak the distributions in ways that are very specific to the purpose they're serving. For example, if it's a database or Web server, people may strip down the base distribution to the point that it's optimized for doing that particular task, but in so doing they end up with a customized flavor of Linux that increases the heterogeneity of the environment - and increases the challenges associated with version control.

Monitoring
While there's an increasing amount of support in the Linux kernel for hardware monitoring, it's important to note that, since Linux is developed by members of the Open Source community, there are different groups of developers at different companies. One group might be working on changes to the kernel for vendor-specific hardware, while other groups might be working on user interfaces. This gives great freedom of choice, but can create challenges for the system manager seeking an integrated, centralized monitoring and management system.

Disaster Recovery Issues
When 10 servers replace a single server, the chance of hardware failure increases more than 10 times. IT managers running Linux on commodity servers realize that the hardware doesn't have as many redundant components and plan for hardware failure. Fortunately, there are a number of high-availability solutions for Linux. Having one Linux server fail in a cluster of 10 servers doesn't impact system uptime, but the system administrator must still reinstall and reconfigure Linux on new replacement hardware. Although Linux on commodity hardware doesn't introduce new installation and provisioning complexities, the move from large servers with redundant components to commodity hardware means that the system administrator is faced with disaster recovery of individual servers on a regular basis.

Patch Management and Deployment
Some enterprise IT managers work in a data center where all their Linux systems are uniform, and they can update all of their machines in one smooth operation. These are the lucky people. I have yet to meet someone this lucky. I generally meet people who are forced to manage some departmental Linux servers that were installed by the "previous Linux guy," customized Linux distributions, and standard Linux distributions from one or more vendors. The great openness and flexibility of Linux can create problems. Enterprise managers are moving to stricter policy enforcement and uniform Linux deployments, but still face significant challenges.

Custom Application Deployment and Package Manager Integration
Most Linux systems rely on a package management system for application installation and deployment. Package management systems are extremely useful for checking for dependencies. Unfortunately, there's generally a lack of in-house expertise in creating Linux software packages. The process can be tedious and many developers find the task tedious. When developing custom enterprise applications, people often install the application without packaging. This creates a significant problem if the new application is broken and the application has to be rolled back to a known working state.

While many IT shops have built "homegrown" approaches to managing these common Linux systems challenges, industry experts see a rapid maturation of enterprise Linux systems management tools on the horizon.

"Linux has moved into the core of the enterprise data center, driven largely by sheer economics," said Bill Weinberg from the Open Source Development Labs (OSDL). "Linux is cheaper than the big iron it replaces, and the hardware and software is cheaper. But for the enterprise data center manager, operational overhead is a key issue - and many analyst firms claim that the cost of managing Linux systems is a barrier to further adoption. We believe that the sophistication of tools for Linux systems management will increase dramatically over the next five years, and further support the enterprise adoption of Linux."