Kevin Gray, product marketing manager for EMC's resource management group, said SCA is a server with discovery software that is agentless, meaning you can get information on SAN devices without installing software on them.

SCA is scheduled to ship in June.

"What they've announced is pretty cool technology for them, in the fact that it's agentless," said Bob Laliberte, an analyst with the Enterprise Strategy Group of Milford, Mass. "One of the biggest pain points for end users deploying a product is that it will have agents. You're seeing EMC coming out with a lot more of these agentless products."

Pricing is based on storage capacity. For example, the 10 Terabyte version costs US$38,480.

SCA is another product in a series that includes Application Discovery Manager, Server Configuration Manager and Voyence, which "does similar things for the network side," Gray said.

Application Discovery Manager, announced in May, 2008, is designed to detect servers, applications and how those applications depend on each other.

Server Configuration Manager, announced last October, discovers details about IT infrastructure and includes compliance toolkits for regulations and standards.

Voyence, which is designed to manage changes to network configurations, is a technology EMC acquired when it acquired a company of the same name in 2007.

"EMC is really putting together a lot of those pieces of the puzzle to give you that overall picture of the data centre," Laliberte said.

SCA is designed to discover hardware in the data centre and detect configuration errors. It also documents the history of changes, providing an audit trail. The software includes templates and wizards designed to make it easier to apply policies to groups of objects.

Gray said it's difficult for IT managers to keep track of items that are either incompatible or incorrectly configured on a storage-area network, especially with more than 1,000 servers maintaining host agents. This in turn leads to network outages, he added.

Laliberte agreed.

"On Day 1, everything is designed correctly, especially with regards to high availability, where you have redundant paths," he said. "Over time, when you've got changes, either planned or unplanned, that's really when the problems start coming up, so in a lot of the cases, you'll get an application outage because a link went down. Typically one link going down shouldn't cause the outage, but because one of the redundant links was taken down some other time and no one noticed it, then you have an outage because you have a failure."