In-Depth

Circling the Data

Active-Circle says it provides a centralized mechanism for establishing policy-based management across your existing network-attached infrastructure. We take a closer look.

• By Jon William Toigo
• 04/04/2006

One of the nice parts of this job is that you get to chat with bright people from start-up companies who are out to fix storage once and for all. Active-Circle, based in Paris, France, is one of these.

Active-Circle has developed software that they regard as true ILM. While I’m not sure I agree with this characterization, I find that their product is a potentially useful tool for managing data across network-connected infrastructure: volumes mounted via the Network File System (NFS), Microsoft’s Common Internet File System (CIFS), or even HTTP/FTP. I recently dove deep into their technology and I see potential merits that might propel its introduction into the U.S. market sometime this year.

Essentially, they're offering a toolset that will allow you bring all of your mounts or shares into a unified infrastructure, whether physically constructed as NAS filers or network-attached file servers. The metaphor is a cellular network in which the cells (individual mounts/shares) are organized into domains. Metadata about the contents of each cell is collected and stored redundantly through the network and is synchronized using a proprietary protocol that runs over TCP/IP.

Active-Circle’s primary claim to fame is that it provides a centralized mechanism for establishing policy-based management across your existing network-attached infrastructure. Similar to some global namespace managers, you can establish target cells where data will be written. You can then construct rules regarding services that will be applied automatically to data in specific cells.

If, for example, Cell 1 contains mission-critical data or data that needs to be replicated for rapid restore or long-term retention, you simply make a rule that copies the data to other designated cells. That way, if Cell 1 ever fails, you can either shift reads and writes to Cell "n" or rapidly restore the identical copies of data on Cell "n" back to a replacement target substituted for the failed Cell 1. You can set this up to occur automatically by writing some fairly succinct policies.

Metadata is replicated, so it would be difficult to lose track of your data or its related policies even if you lost one of the metadata servers deployed throughout the infrastructure (white box servers that operate Linux for now, Windows soon).

It isn’t Information Lifecycle Management, by the way. The heavy lifting of building a data classification scheme will still require the consumer to do significant upfront analysis. Moreover, there is no storage classification scheme that differentiates target platforms under the cell metaphor based on their price or performance: you need to create that, too. Furthermore, there is no access frequency counting function that I can see, which would give you a granular way to determine when data might be migrated to an archive based on usage. Without these features, I can’t see how you can call Active-Circle (or any other product representing itself as ILM) “true ILM.” But what Active-Circle does do, it does well and elegantly.

My advice to the company, in its efforts to bring the product to the U.S, market, is to dumb down its messaging about ILM because no one believes it, and because they are even less likely to believe it coming from an unfamiliar vendor. If EMC couldn’t get anywhere with its ILM palaver, what chance does a small company in Paris have to boil the storage ocean?

An immediate value of the solution might be as a high-granularity replacement for snapshot-based protection for Exchange Mail and IBM Lotus Notes. The latter messaging application, still in widespread use, is desperately lacking in terms of data protection and rapid recovery. Most messaging protection solutions, in fact, depend on snapshots that provide little or no granularity of information about the messages themselves. Restore is all or nothing for the most part.

Active-Circle’s metadata repository could change that. You could do selective restore, as I see it, of all messages prior to the culprit message that fouled up the .PST file or Notes database. The product collects all of the necessary versioning information if you ever want to do selective restore.

You could also readily extend the functionality of this product to do selective deletes across all instances of data in your production storage environment. I suspect that you would be able to define a policy that states that a certain class of file should be zapped when its age exceeds the required holding time specified in regulations and laws. That would be a good for companies that have been busy figuring out how to retain data for compliance that they have given little thought to how they would delete it at the end of the retention period.

Active-Circle has a lot of bells and whistles that you would want from a product like this. It does encryption and other voodoo functions you might need—on demand and in a sane context. The interface, like the Web site, is still only available in French, but it will be available in English soon for those of us who do not parlez-vous.

Working on a niche market first would be the safe play for Active-Circle—they might want to characterize their product as a better way to replicate and protect messaging, and to delete data that has outlived its required retention period. They have some cross-pollination with Atempo, by the way, so perhaps they could extend this story to include not just data on production storage, but on tape as well.

Working out from this niche, the company might be able to convince Global 2000 companies to use their technology as a component of a much broader scheme for data management. From where I’m sitting, the product is worth a look.

Your comments are welcome: jtoigo@toigopartners.com.