It is refreshing to see some within EMC finally becoming concerned about integrated information management.  Chuck Hollis posted a very nice blog entry today.  I agree with many of Chuck’s points about how important information has become and, yet, how limited we are due to the complexities that the information systems we store it on create.

In his blog, he mentioned a few things that I wanted to provide some of my own insight on. 

The CFO of Information:

Mr. Hollis mentions this concept of a “CFO of Information” whose job would be to manage corporate information much like a CFO of today would manage corporate finances.  What a difficult job indeed! Unlike numbers, metadata comes in many forms, types and sizes which makes it largely incompatible to easily interpret between multiple complex systems.  Where a CFO’s job might be to add 1 + 3 + 9, a CFO of Information’s job would be more like adding 1 + “a square” + “a banana”.  You would have to be on a Netflix commercial to get that right!

Chuck rightly hinted that EMC’s large and segmented product portfolio has the potential to create massive amounts of incompatible metadata.

“If you take a tour across the EMC porfolio, and ask the question "what parts of the EMC offering generate and store metadata", and "what products can use metadata to implement policy", it’d be a very long and detailed answer.”

When you consider information management in an EMC shop (or practically any IT shop), there are many, many metadata repositories. If you backup your virtual data, EMC will store that metadata in one repository (Avamar).  If you backup non-virtualized data, EMC will store that in a separate repository (NetWorker).  If you need replication of that backup data, that metadata gets stored in RecoverPoint.  If you are doing archiving, there is metadata for DiskXtender, SourceOne.  For content management, there is metadata for Documentum. And on and on the examples go for eDiscovery, Compliance, SRM, etc…


Now, Chuck normally doesn’t present problems that he doesn’t already have solutions for (I was actually surprised to see no solution in his post!), so I suspect EMC is well on its way with some kind of offering here. Possibly their approach will be to provide a more centralized tool that gleans information off a few of these other metadata repositories to create new metadata in another repository. In this instance, I suppose the information about the metadata would be considered meta-metadata. For the sake of science and continuation of the species, I hope they don’t accidentally mix metadata and meta-metadata in their labs!

Regardless of what direction they take, it is going to be ugly.  Developing a framework that can add numbers with shapes with fruit and come up with some legible result is difficult enough—to do this in a way so that the information is so reliable you can make business decisions based on it is something else entirely.  Especially, when no one underneath the CFO of Information can actually explain to him or her how the EMC Black Box of Anti-Mattered Meta-MetaData comes up with the information it does.

A Better Way?image

What if EMC actually integrated their products into a single data management platform. How much easier would it be if they had a single product that stored metadata for all types of backups, all kinds of archiving, replication, eDiscovery, SRM, compliance, etc.. in a single location.  What if it also included a centralized policy engine that allowed for the control of that data for protection and governance needs as well as to dynamically move that data—as IT needs change—through multiple tiers of storage and even to the cloud and back.  The benefits would be incredible.  Instead of dozens of management frameworks and policy engines, a customer could have one data management platform, one policy engine for control and a single metadata repository where enterprise wide information intelligence could be obtained that is explainable and rock solid enough to make business decisions off of.  Now that is what I think customers are looking for!

The great news is, this technology is already available with  Commvault Simpana.

Leave a Reply




You can use these HTML tags

<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>

No CAPTCHA challenge required.