EMC? Server Management?

In the "I'm surprised that people are surprised" category for this week: EMC's recent announcement of SCM (Server Configuration Manager) and a higher level tool (Configuration Analytics Manager).

More than a few journalists were surprised to see EMC having an offering in this space.

I guess we're just not doing a good job about getting our strategy out there -- because the announcement is a natural progression of what we're doing in the broader category.

So, let me try to explain?

Targeting Management Efficiency

In just about every IT survey you see, you'll find different buckets: hardware, software, staff and so on.  And it's hard to miss that a big hunk of just about every IT budget is the people who keep the lights on for the business, so to speak.

Lots of attention on keeping service levels high, and responding to new requirements by the business.  Just so long as you can do a better job with fewer people, everyone will be happy.

Several years ago, EMC decided that it wanted to be a player in the broader category of "resource management" -- the task of dynamically orchestrating IT resources to meet business needs.

Sure, hardware was becoming more efficient -- and is becoming even more so.  But how do we make people more efficient?  And we believed that several industry trends would force IT shops to rethink this whole topic in the next few years.

Clearly, the established players weren't making headway.  Lots of management software was being sold by various vendors, but the basic industry averages (e.g. headcount per application, cycle time to process an IT event, etc.) were stubbornly refusing to trend downwards at the same rate that hardware efficiencies were.

Was this a problem, an opportunity, or both?

The Big Idea

Several years ago, EMC saw that IT needed a model to run their business: one logically consistent representation of physical views, logical views, application views and so on.  Focus on the logical relationships between entities, rather than providing simple inventories.

Then, use this model as the central integration point for all the things that IT cares about: discovery, root-cause analysis, ITIL workflows, IT compliance, domain management, and -- well -- server configuration management.

And, over the last few years, we've been predictably marching down this path. 

It's starting to work -- but it's been a long road so far.  And we're not done yet, but we've made great progress, IMHO.

So, what's this all about?

Everyone Wants Their Own Tool

Predictably, every discipline in IT wants their own set of tools.  Network people want theirs, application people want theirs, help desk people want theirs, and so on.

Great for all the individual parts of IT, but where does it all tie together?  How can everyone get to a common view of "truth", and understand how their bits relate to other bits?

I remember many years ago, I went to a budget planning meeting.  After about 30 minutes, it dawned on us that everyone had a slightly different take on the numbers. 

That was problem #1.

The second problem is that we had very limited knowledge about how decisions by one group to invest or not affected others in the business.  We had no model for interaction.

Job #1 was creating a consistent view of expenses, forecasts and budgets.  That much was "easy", relatively speaking.  Where we really lost time -- months in fact -- was coming up with a model on how the activities of one group affected other groups, and the business as a whole.

That was downright hard for everyone.  It didn't really matter that much what each individual group thought was "best", we had to take a larger view that was a stretch for all involved.

True in the corporate budgeting world, also true in the IT world.

Narrow Vs. Wide Views

If you look at the SCM offering from EMC, it's fair to stack it up against similar products that claim to do similar things.  Sure, it does things like test for compliance with configurations and patches, does server and desktop patch management (including virtualized ones, natch), and provides remediation capabilities.

But that would be the more narrow view, IMHO.

The wider view is the ecosystem that it plays with.  For example, the ability to have Smarts ADM dynamically discover configuration details regarding applications, servers, networks, storage, etc. -- and their relationships, dynamically populating the Smarts model that SCM uses.

Or the fact that it works hand in hand with VoyenceControl's network discovery and compliance management capabilities, so you can seamlessly work between the server domain and the network domain -- again, using the underlying Smarts model.  Or that it works with Infra's web-based ITIL automation tools, so that workflows can be driven from Infra into SCM, and potentially back again.

Or that Smarts' real-time root cause analysis capabilities work with the same model.

There's more to the story, but I think you get the gist of what's cool here -- there's magic in the model.

Triggers To Rethinking Resource Management

There are a couple of secular trends in the industry that I think are causing more and more IT organizations to re-think their overall strategy towards end-to-end IT management.

The first one has to be the continual decomposition of applications, services and resources into multiple pieces; all of which need to be managed and orchestrated as a whole. 

No longer can we walk up to a single server and say "there's the application that the users are using".

The second one has to be the immediate and sharp focus everyone seems to have these days in dramatically reducing costs and improving efficiency while continuing to provide high service levels, be compliant, etc. 

Newer approaches to resource management and IT orchestration just have to be part of this story.

And, finally, there's no denying the wholesale shift to virtualized environments where things are inherently dynamic and fluid. 

Tools that were designed for a static, partitioned view just aren't holding up in these newer virtualized and converged environment.

And What About Leadership?

One especially petulant taunt came from someone at HP, who wondered loudly how EMC could claim to be a leader in this category.

Well, certainly EMC can't claim leadership in terms of market share.  Nor can we point to tens of thousands of installations, unless of course we're talking things like storage management, or network management, or aspects of security  management, and so on.

No, those claims go to integrated management vendors who built their capabilities in a time long before virtualization, and deconstructed IT applications, and integrated use of external service providers, and IT compliance, and ... well, long before the emergence of the most interesting facets of the current IT landscape.

The real question should be -- who will be the next leader?

Courteous comments welcome as always!

 

Comments

[Martin G]

I kind of wish that EMC would concentrate on making ECC work properly; make it scale, reliable, produce the reports I need without engaging EMC PS etc. Make what you've got work first please! Give us the rewrite which has been promised so often and then move onto doing server management. If any of my server colleagues ask whether taking a tool from EMC to manage their servers is a good idea, I'm afraid that the answer would not be positive today. Sorry to be so harsh but if you heard some of conversations I've had over the past few days...

[Stuart S]

EMC have been trying to play in the "management space" for some time, and IMHO have not managed to get mangement of their own core product set right(read storage products) before they go chasing after other products to manage. Please please get your own house in order and get the basics right before trying to intergrate every other piece of infrastruture

To be more exact, and talking about a specific product - ECC, it has such an extremely poor name in enterprise infrastructures, does not scale, does not manage storage effectively and in lots of cases requires dedicated staff to manage the management product (that just reads plain wrong doesnt it...)

That being said, there are a number of core products that EMC own (SMARTS ADM etc) that could give fantastic holistic views, and if they could be tied together (with a light / easy to use application) - good stuff!

I do agree with the article in general, and a holistic view into infrastructure is the way forwards (especially in the new shiny virtualised world), however I refer back to my top comment - please lets get the basics right first, and be able to manage them in a manner that is both scalable and managable (i.e you dont have to be resource intensive to manage the management layer).

- Basic Element managers
- Tie them together with a "light" management layer (Technology specific - i.e. storage)
- Wrap a manger of mangers over the top for holistic view

please please please fix ECC and DO NOT TRY AND BOIL THE OCEAN...

[Chuck Hollis]

Thanks, Stuart, message received LOUD AND CLEAR.

I do a fair amount of interaction with the group at EMC who does management products, and I would offer that they share many of your views and perspectives.

The best part? They're making the investments required to do exactly what you suggest. Unfortunately, it's taking a bit longer than anyone would like.

Thanks for commenting!

[Calvin Zito]

Chuck,

Since you misrepresented my "petulant taunt", I'll repeat it here. I didn't say EMC was claiming leadership with your new server management software. I said I find it ironic that you’re making some weak arguments why server/storage vendors don't make good storage vendors while EMC is trying to become more than a storage-only vendor by announcing server management software.

I pointed out the hypocrisy of your arguements. Maybe you should spend more time fixing ECC and less on self-serving attacks.

Have a good weekend - the Petulant One.