AutoVirt followed other file virtualization companies and products to the grave. Industry experts say IT managers have a hard time justifying its purchase to higher-level executives as a need-to-have rather than a nice-to-have tool.
“I’m not surprised,” Marc Staimer, president of Dragon Slayer Consulting, said of AutoVirt’s demise. “AutoVirt was not solving an urgent problem. It was an important problem. People tend to spend money on urgent problems and put off spending money on important problems.”
The Nashua, N.H.-based AutoVirt launched in 2007 with its file virtualization software, which helped businesses analyze data stored on Microsoft-based file servers and any NAS hardware. The software used Microsoft APIs to integrate a proprietary global namespace with Microsoft’s Domain Name System (DNS) server. AutoVirt offered three editions of the product, called AutoAssess, AutoManage and AutoMigrate.
AutoVirt raised about $20 million in funding over the years but failed to come up with another round, which it needed to survive.
AutoVirt’s targeting of the SMB and midmarket level contributed to its demise, said Steve Duplessie, senior analyst at Enterprise Strategy Group (ESG). The software only worked for Windows environments in smaller organizations, which typically have small amounts of data to migrate, compared with enterprise-level customers that need to move terabytes of data. Duplessie said the main perceived benefit of AutoVirt for SMBs was that it saved the IT manager time by automating migrations.
“It was a great product. They automated the data migration process for Windows machines,” Duplessie said. “But the IT guy had to justify to his boss that the benefit is he would have more free time. There was no direct benefit to the company.
“AutoVirt never armed their customers with enough ammunition to fight the internal battle. It was an execution failure on the business side, and it didn’t look like it was going to hit critical mass so the venture capitalists gave up.”
AutoVirt may have had a better chance if its software also supported Linux and Unix because that would have brought it into the enterprise, Duplessie said. The average data migration time for a small organization with a Windows environment is about a day, while a larger company could require eight to nine months from planning to execution.
“They didn’t support a Unix environment, where the problem is more expensive and a bigger light is shined on that problem,” Duplessie said.
AutoVirt positioned itself as an alternative to Microsoft’s Distributed File System (DFS), which is delivered as a free feature.
The other roadblock AutoVirt’s file virtualization appliance encountered is it primarily sits in front of a CIFS network-attached storage device. That can cause support problems.
“The problem with appliance-based global namespace -- and it’s a problem with every storage appliance -- is you put the appliance in front of another vendor and it causes root-cause analysis problems,” said Staimer. “The support organizations wind up pointing fingers at each other.”
Several years ago, file virtualization was considered a hot, emerging technology that would help with NAS sprawl. The initial market goal was to provide scale-out management of multiple traditional NAS systems by layering a third-party global namespace over existing NAS nodes. But many of the vendors have folded, and others have repositioned their products.
NeoPath Networks and Attune Networks are gone. Brocade Communications Systems Inc. discontinued the StorageX file virtualization platform based on its 2006 acquisition of NuView Systems Inc. EMC Corp. refocused its Rainfinity file virtualization product as a data migration add-on for tiered storage and to move data to the cloud.
F5 Networks Inc.’s ARX platform and Avere Systems Inc.’s FXT appliances are still standing. F5 acquired file virtualization startup Acopia in 2007. Although F5 has continued to develop and push its file virtualization, ARX products accounted for only $7.3 million in revenue in the last quarter of 2011 -- less than 4 percent of F5’s overall product revenue.
Avere has had success using virtualization and caching to accelerate NAS systems. “Avere is a different play,” Staimer said.
Staimer said the plight of file virtualization is part of a larger problem with storage virtualization in general.
“Storage virtualization is a feature, not a product,” he said. “It’s tough to sell a feature as a product.”