There’s no bigger source of Big Data than machine-generated log files. And every business has them. That means there’s a huge opportunity for vendors that monitor and analyze log file data to help developers improve application performance. Just ask Splunk, which is reportedly planning an IPO in the near future. Some industry watchers value the company at $1 billion or more.

Loggly, a San Francisco-based start-up co-founded by ex-Splunker Kord Campbell, also specializes in monitoring and analyzing machine-generated data. But, unlike Splunk, Loggly offers its software as a service (SaaS) rather than as an on-premise download. Kord told me Loggly’s SaaS model removes significant barriers to entry and makes it easier for developers to get started with log file data monitoring and analysis.

“While we were at Splunk” – about a third of Loggly’s 15 employees are ex-Splunkers – “we were aware of what the limitations were of the software sales cycle. So you need to get somebody to download it, you need to be able to install it on a server, they have to maintain and trouble shoot it on their own,” Campbell explained.

Loggly customers don’t have to worry about configuring and maintaining related hardware or performing software updates. Instead, customers sign up for a free 30-day account, then “simply configure their syslog install, whatever that may be, to send us their log file data,” Campbell explained.

Loggly takes it from there. The company time-stamps each log file, then applies a full text index to the data to make it searchable. After the data is loaded into Loggly’s search engine, which is based on the open source Solr Lucene search platform, customers can search and otherwise segment their log files to identify patterns or highlight exceptions.

“So if there’s a 404 error, we catch it and index it by pulling out text,” Campbell said. “We can then show [the customer] graphs of that term over time in their log files.” The resulting analysis is stored in MongoDB, the document-oriented open source NoSQL database.

Kord Campbell, Cofounder and CEO, Loggly

After the 30-day free trial is up, Loggly offers users a monthly paid subscription based on their usage during the previous month. Those customers that opt not to sign on as paying customers can still use the service, but with scale restrictions and less support. Since its launch in February 2011, Campbell said about 5,200 users have signed on to the free 30-day trial, with about 250 converted to paid accounts varying from $100 to $400 a month in subscription fees.

Loggly’s paying customers include Salesforce.com, the Washington Post, and Electronic Arts, but I think the company’s SaaS model could prove particularly attractive to start-up web application vendors for whom app performance is often key to spurring adoption. They are often small shops that lack the capital needed to invest in expensive on-premise software, so an affordable SaaS-based option could be just the ticket they’re looking for. Campbell said Loggly is already seeing significant demand from these types of vendors. “Anybody that runs a web app, they’re calling us,” Campbell said.

The SaaS model presents some challenges, however, namely around the amount of data that Loggly can ingest over the public internet and store efficiently. Campbell said Loggly has a handful of customers that send the company between 100 to 300 gigabytes of log file data per day, but that it is working to scale that number to closer to half a terabyte.

Loggly, which has raised close to $5 million in VC over the last two years led by Trinity Ventures and True Ventures, is also working to increase the amount of data it can store per user. “This quarter we increased the amount of data we’re able to process on a box by ten-fold,” Campbell said.

Campbell said his engineers are also in the process of building applications that sit on top of the Loggly platform for specific functions that can be easily integrated by third-party service providers. Loggly already offers an alert application that notifies developers when an application is behaving unexpectedly, and is also working on performance monitoring and exception reporting applications that Campbell expects will be released later this year.

“We expect that most of the SaaS revenue that we will be generating will come in off of third party applications moving forward,” Campbell said, noting that the Loggly platform was built on standardized web technologies to encourage just that. “But you can’t build a platform over night.”