Copan Systems announced a new archival version of its massive array of idle disk (MAID) technology that allows users to retrieve their own files from online in a way that prevents them from corrupting existing files.
The hardware, which offers retrieval of data based on metadata and not hash keys, is aimed squarely at taking market share from EMC's Centera content-addressed storage array.
"This thing puts [Copan] into the archiving game with a much better price/performance," said Arun Taneja, founder of research firm Teneja Group in Hopkinton, Mass.
The new array, the Revolution 200A, was announced at Storage Decisions and is based on the company's existing hardware platform. The hardware will be widely available in January and will sell for about US$3 per gigabyte of capacity, according to CEO Roger Archibald.
Unlike its predecessor, the Revolution 200T, which emulated a tape library to backup servers, the upgraded model allows IT administrators to store data in file format for long-term archival.
Peter Carucci, a technology consultant who works for the New York Stock Exchange (NYSE), said he could see the NYSE's IT department adopting the technology because of its ability to archive online hundreds of terabytes of data in a single rack. "We're definitely going to look at it," he said.
Copan's Revolution 200T is an array filled with relatively inexpensive ATA-based disk drives that can scale up to 224TB of capacity and has been positioned as a virtual tape library by using software from FalconStor Software.
While the Revolution 200T sat on a storage-area network, the 200A model is designed to sit on a LAN and has four 1GB Ethernet ports.
A new software application called Millennia Archive stores data on the 200A using the SHA-256 hashing algorithm to ensure that each file is unchangeable. Any change to a file automatically creates another version. The array also allows users to search for files by name, data or other characteristics stored in metadata.
"Application servers can either schedule copies of data [to other arrays] or the migration of data," said Archibald.
Taneja said Copan has an advantage over EMC's Centera in that it uses SHA-256 instead of the MD5 hashing algorithm, which is far less secure. "The problem with 128-bit MD5 is it's possible to have two files that give you the same hash [index]. SHA-256 is not foolproof, but it's two to the power of 128 more secure," he said.
Another advantage Copan's array has over Centera is that it can store and more quickly retrieve many more files than Centera -- up to 1 billion files on a single array. "That's significantly higher than Centera," Taneja said.
Copan claims its arrays can quadruple the life of disk drives by shutting them down when they're not in use. And because most of the disk isn't spinning, less heat is generated and more storage hardware can be used in a smaller space.
Craig Taylor, associate director of open systems at the Chicago Mercantile Exchange, said the exchange currently owns a Revolution 200T and has been waiting for Copan to create a write-once, read-many version of the array for regulatory purposes.
Taylor said that over the past year the exchange has used the 200T, he has not lost a single disk drive. But he said that he has been waiting for "a compliance front-end on our box," which he will have with the 200A.
