Startup Zmanda Inc. came out of stealth this week with the launch of an open source backup product based on the Amanda open source backup and recovery software.
According to users and industry experts, the features and support Zmanda has added to the package could be appealing to highly cost-conscious but less technically advanced users at the low end of the market. These users are currently caught between writing their own shell scripts and using freeware or enterprise-level backup software they can't afford to buy or manage.
The Zmanda Network with Amanda Enterprise Edition software is based on the Amanda standard developed at the University of Maryland over a decade ago. Zmanda has developed it from its original incarnation to support Windows machines, added security features, brought it up to date with support for new applications, improved its scalability and added a management GUI.
Some users are already using a straight-up version of Amanda, but in a way are also using Zmanda's product, as "just about every developer who has really worked on this standard in the last few years has done it through us," according to Zmanda's CEO, Chander Kant. For example, Zmanda has addressed some of the scalability issues in the original Amanda specification, one of which was that it did not allow a single backup to span multiple tapes. Shops using older Amanda versions had to run separate different configuration scripts for each individual tape.
The company's newest product, however, is expected to bring open source backup software into a wider marketplace -- a marketplace that, according to backup expert Curtis Preston, senior analyst with GlassHouse Technologies Inc., has been underserved.
"I do see a chasm between the free or $100 products like the scaled down versions of software that come with tape drives and the products that cost tens of thousands of dollars," he said.
Companies that can't afford a name brand backup product but have needs beyond freeware often write their own shell scripts, a labor-intensive task Preston describes as "onerous. I wouldn't wish it on anybody."
The Zmanda service, which costs between $50 and $250 a year depending on level of support, will find "a strong market," according to Preston. "For every Exxon, there are thousands of small shops," he said. "From that you can get a sense of the market for this type of product."
The ideal shop for Zmanda's product looks very much like the one overseen by Charles Smith, systems administrator for CyCorp Inc., which conducts artificial intelligence research and relies on government subsidies for funding. Smith said the company has been using Amanda to back up eight Linux file servers, three Windows servers and about 40 Windows PCs, all with DAS. He recently updated to the newer version developed by Zmanda. He said he hasn't decided yet whether to sign up for Zmanda's subscription or implement its GUI, but said it could mean Amanda has a much broader reach than it has in the past.
"A lot of users out there are only familiar with Windows software that has GUIs and aren't going to understand or be able to work with Amanda," he said. "It won't be terribly useful for someone who's not used to that type of thing."
Smith said he had already stopped using Veritas Software Corp.'s Backup Exec product, and called it a "terribly buggy" product. But, he said, his end users still miss a feature Backup Exec offered that would allow them to do their own restores.
"If Zmanda can bring that type of functionality to Amanda, that would definitely justify the extra costs," he said.
"I think it's definitely time open source came out of the closet, so to speak," Preston said. "The good news is, the open source backup community has come out with some really strong products in the last few years."
Preston cited other open source backup tools called Backup PC and Bacula that had been developed in response to the limitations of Amanda "because before Zmanda got involved, nobody thought they would ever get fixed," Preston said. Those products, too, already have an install base in the thousands and could also make an impact, he said, citing other open source companies, like Red Hat Inc., as an example of how open source companies can succeed.
"Everyone thought they were stupid when they first got started," Preston said. "Nobody understood how they were going to make money selling free software. But I think they did OK."