Analyzing Streaming Media Server Log Files
Sawmill includes rich documentation that explains how to install, operate, and customize the software. Sawmill can run either as a CGI program, or as its own standalone web server if you don’t have a web server with CGI access. Sawmill runs on practically every OS and has a simple install process. After you install the software, you create profiles to process your log files.
Sawmill is very, very slick. From the install process through profile creation and report generation, it’s rare that you need to consult the documentation—it just works. For example, creating a profile couldn’t be simpler. From the Administrative interface, you click the Create New Profile link at the top of the page. This causes a small window to pop up, which then walks you through the process. First you tell Sawmill where to find the log files you want to process. If the log files are on the same machine, you can browse to the directory using the Browse button. If they’re on another machine you can tell Sawmill to retrieve them using FTP or HTTP.
After you’ve set your log file source, Sawmill analyzes the files and attempts to figure out what kind of log files they are. You then confirm the type of log file, or choose another log format if Sawmill misidentifies the format. Next, you select which fields in the log file you want to process, and finally you choose a name for your profile. When you’re done, Sawmill processes the log files and drops you on the reports page. It’s that simple.
Sawmill has navigation on the left that allows you to extract rich detail from your log files. There are also icons along the top that allow you to filter the data by date, or by just about any other condition and by any field. This is where the real power of Sawmill comes in. You can slice and dice your information as many ways as you’d like, and then export it in a CSV format if you want to import it to another application like a spreadsheet.
In fact, this filtering can be applied when Sawmill processes your log files. So, for example, if you have an account with a CDN, the URL field of your log file entries may include a number of directories specific to your CDN that you don’t really care about. You can create a filter to filter them out. From the administrative interface, click on your profile, then select Log Filters from under the Log Data menu. Click New Log Filter, and Sawmill brings up a wizard that walks you through the process. You can create filters that are dependent on a condition, filters that simply perform an action, or special advanced filters using their scripting language. You can apply multiple filters to a profile, and arrange them in any order.
The look and feel of Sawmill is also highly configurable (except when using Sawmill Lite). You can control how reports look, and how data is displayed. You can control which reports are available for each profile. Sawmill also includes a built-in user database, so you can create unlimited user accounts and associate them with different profiles. You can give users administrative privileges for the profiles they’re associated with, or reserve that function for yourself.
Finally, Sawmill provides a rich set of system administrative functions that allow you to schedule tasks, such as updating the stats every night. You can even schedule regularly emailed reports for your clients.
Sawmill provides multiple layers of support. There’s a Sawmill FAQ and a very active forum. If you have questions or concerns, you can email support, which promises a 24-hour turnaround. If you want high-touch treatment, you can buy premium support for an additional 25% on top of your licensing fee, which gets you a dedicated email contact address and phone support, as well as free updates.
Conclusions
So there you have it. Thousands upon thousands of streaming media servers out there, and these are two of the only viable stats programs to choose from. AWStats is a little light on features, and it’s not meant for folks who aren’t comfortable installing software on servers via the command line. But if you’re on a budget and interested in finding out which of your files are most popular, it’s a great option.Sawmill is a fully featured system with the power to satisfy any of your log-processing needs. It will slice, dice, and julienne your data any way you want, and let you control who gets to look at it and what they see. It’s not free, but if you’re trying to decide whether to build versus license, you’d be hard pressed to build something as good as Sawmill for the price of the license.
Full disclosure: After doing the research for this article, I decided to license Sawmill, and haven’t had a moment of buyer’s remorse.