When the word “systems management” comes up, it’s easy to think the discusion is about “enterprise systems management.” That’s because that’s all there’s been for a long, long time. Indeed, looking at it, the notion of tacking “enterprise” in front of “systems management” seems redundant. It’s been all enterprise all the time for a long time. Sure, there’s been small network and “ping” monitors, and plenty of mid-market platforms like my favorite title for a monitoring app: WhatsUp. I always want to say “WhatsUp, home-fry-skillet-pan? How’s that router shakin’?!”.
(I should probably clarify for those of you don’t know me personally that I separate out being able to say something funny about a word from making fun of what that word represents. I am a multi-layered smart-ass.)
Long-tail Systems Management
Of late, the enterprise systems management crew has been questing for the suite spot (pun intended ;>) to sell to the mid-market. The target sales are people who don’t want to sign an “enterprise license” and/or gorge on the stream of software and configuration that comes with enterprise systems management. The arguments that we hear from the OSS systems management vendors, closed source people like RTO or CITTIO, and now Microsoft tend to start with the line: “now, we all know that [enterprise] systems management platforms are way too complex and have too much in them. Not everyone needs that…”
I should side-note that I don’t fully buy into that complexity pitch. The reality is much more complex (rich irony!). Most the systems management offerings out there can be complex, and some of it should be. But as someone who used to be on the front-lines of the fight, I can tell you that there’s a constant internal battle to make things as simple as possible. Furthermore, despite my love for Conway’s Law, the size of a company doesn’t always drive the complexity or simplicity of the software it produces. Take Nagios configuration: it’s way too much for simpletons like me.
Dubious complexity discussions aside, the field of mid-market/long-tail/SMB systems management is wide open. From what I saw at MMS 2006, Microsoft is a powerful contender. And though the the low-level technical sessions were the only place where the word “open source” came up (as we’ll find out, we should move that discussion up the stack all the way to the keynotes), the fact of the matter is that Open Source systems management and Microsoft systems management are going to be going at it head-to-head over the next few years. Both of them are targeting the same market.
OSS Systems Management
From the discusions I’ve had with OSS systems management vendors, they seem to be as oblivious to the Microsoft threat as Microsoft is of them. This might be because I haven’t thought, until now, to ask the question, “what are you going to do about Microsoft?” But, for all the railing each of them (and the closed-source mid-markets) does against the big 4 — BMC, CA, IBM, and HP — they’ve clearly spent time thinking about who they’re up against.
Worrying about the big four at this point is a distraction. The OSS systems management vendors need to start thinking about Microsoft’s management and CCM offerings: the newly named System Center suite. Sure, the roadmap is long, but MOM and SMS are here today, and they’re quickly combining their powers, as it were, into System Center. The possibility of more Vista biffs aside, 2007 will come quicker than we realize.
Strange Bed-fellows
Microsoft can and does do well in the mid-market. More importantly, they have something to offer now and cheaply (MOM and SMS) that has the potential for a bright future. In that sense, mid-market systems management is the classic battle of Windows vs. *nix one. The more Windows in the mid-market, the easier it’ll be for Microsoft, the more *nix, the easier it’ll be for the OSS crew.
To me, this means that right now is a great time for Microsoft and the OSS systems management crews to start working, or at least “partnering” closely together. Never mind what the public messaging is on such a relationship. The real reason is pure samurai mind-games: “keep your friends close, and your enemies closer.”
For Microsoft the advantage is keeping an eye on the innovation that’s occurring in open source systems management. For the open source systems management folks, the advantage is cracking into the Windows silo. Microsoft’s interest may not seem as great, but the fact of the matter is that the majority of the innovation occurring in systems management is coming from the open source companies and projects.
Larry Ellison’s views on OSS IP aside — Good luck with that! — OSS systems management people are the ones who’ll be, and are, delivering innovation in the area: hosted systems management, collaborative systems management, RSS, simplicity, new approaches, and perhaps even new service models around systems management.
Let’s digress down that rabbit hole a bit…
Open Source Systems Management Driving Innovation
I’ll save a lengthier discussion of innovations in OSS systems management for later, but here’s a little taste:
- FiveRuns is just now unwrapping their hosted systems management offering, whose UI is built on Rails of all things. I’ve spoken with Dave Wilby at length about all sorts of exciting (but sadly, for this post, embargoed) features, that I keep waiting to see in action and talk about.
- Qlusters is lucky to have CTO William “whurley” Hurley driving their quest to, frankly, take over and revolutionize systems management. Lofty goals, sure, but not impossible, esp. with the foundation they have available today.
- GroundWork is wrapping up OSS systems management tools into digestible packs and wrapping service sales around that stack. From the conversation I had with them this week, they also seem to have a very exacting understanding of how they’ll make money and the chops to deliver.
- Along with closed-source members LogLogic and LogRhythm, the Log Management and Analysis crew includes OSS-based Splunk, who’s taken a great crack at a new approach to systems management. More importantly, Splunk Base is one of the first attempts at true collaborative systems management.
While I have to disclaim that I haven’t talked with customers for any of the above yet (meaning you could justifiably dismiss it all as slide-ware and cooked demo’s until I do talk with actual end-users), what I’ve seen is impressive. What’s even more impressive is the passion, in true open source form, each of the companies have for their work.
More importantly, those are just 4 companies that to talked with (please email me if you think I’ve missed you ;>). There are many other companies, projects, and people out there innovating in the seemingly “innovation dead” field of systems management. For example, some day soon a mad midnight coder (or one of the above) is going to figure out how RSS can clean up the data avalanche…or do something totally unexpected with mashing Web 2.0 concepts into systems management. Perhaps it’s already happened.
What the OSS systems management people don’t have enough of is a the lust for ITIL-based thinking that the big 4, and even Microsoft, are focused on now; GroundWork may be an exception to this. I like the structure and common language that ITIL brings, so I of course want to see the OSS crew pick it up more. That’s something else they could learn from a closer relationship with Microsoft, or the big 4.
Working Together at Arm’s Length: WS-Management and SDM
One mutually beneficial area to work on are systems management standards. I can’t speak to the technical merits of WS-Management (I’d love input on that), but it’s something that’s there to consider. Microsoft is bullish on it, and they’re back-porting it to “older” Windows (someday, at least ;>)
I’m eager to see a successful common data format emerge as well. Many have tried and died. But, having programmed at the lowest levels of systems management, I can tell you that there’s nothing special or differentiating about the data model you use. It’s all just different words for the same thing. Pure Carr.
While Microsoft has the System Definition Model (SDM), they’re keeping it close to their chest and tied up in licensing. I have a feeling that, SDM might be too heavy weight because of that. When I asked them about standardizing SDM, Microsoft folks focused on the bad parts of committee driven formats to justify keeping SDM closed. But they’re missing the good parts of community driven formats; one of them being that those formats tend to be light weight.
The OSS systems management camp could help with that right out of the gate. More importantly, if the OSS folks added support for those standards, the relevance of System Center would grow because it could pull even more data from outside the Windows silo and push management outwards as well.
I’ll Take the Ecosystem I can Get
Ideally, Microsoft and the OSS systems management folks would genuinly work together and create a healthy ecosystem that could move it’s attention fully to solving everyone’s systems management problems, instead of focusing on the zero-sum, market-grab, logic they’ll really operate under. But, come one: software companies and groups never get along. They always come in at least pairs: WS-Management and WSDM; Java and .Net; Rails and Enterprise Software; Redmond and Armonk.
Despite that, Microsoft and the OSS systems management folks could still benefit from a continuous pairing: if only to smell each other out for the cold war, as it were. Continuing with the WWII metaphors, the two might be interesting allies against the big 4: it’d be the embattled elders calling the plucky youngsters from across the pond.
Sun, of course, is in an interesting position in this scenario. They’re the yet to be played wild-card in systems management at the moment: ever on the verge of open sourcing N1. Indeed, they could replace Microsoft in the above, though in a different way, no doubt. Perhaps they should jump on the chance before Microsoft has the time to turn the fleet around ;>
Disclaimer: Microsoft, Sun, BMC, IBM, and LogLogic are client.
Technorati Tags: bmc, conferences, enterprisesoftware, ibm, itil, microsoft, microsoftmanagementsummit, mms2006, systemsmangement, ws-management
Coté you have freaked me out with common interests yet again on that reference to The Book of Five Rings – that book is on my bedstand! Way too weird… but it’s a very good read. Not to mention that you sound way more hip (== awesome points) quoting that than Sun Tzu. 😉
Great post!
Does flexibility and openness mean complexity? How much should vendors contribute out of the box versus allowing customers to implement what works for them?
I’m all for both – flexibility, openness and contributed content (from the vendor, partners and community of users). This is where I hope to make a difference.
The thing I always heard about hurting the most were the rogue groups with their own OSS tools outside the “group” who was supposed to provide systems management services. This leads to even more finger pointing with the NOC or support group when “my tool says this” and “our tool says that”. Well integrated into IT Operations and the business, OSS systems management have a place for sure and can serve as a stepping stone into other solutions as the needs of the business change.
Do you share similar thoughts on Cisco as you do for Microsoft? Do you see and OSS systems management groups climbing the stack into the process and business (BSM/BAM/BPM) areas?
Doug
BSM/ITSM Blog: http://dougmcclure.net
Doug: first apologies for taking so long to respond.
On the other hand, that delay has brought about an interesting answer to the finger pointing question. Ideally, we'd want some way to tell what's wrong in a system regardless of which tool was used. Put another way, we'd want to get the same conclusion about what's wrong regardless of which tool is used.
To my mind, these means we need something more low-level than ITIL as a standard. A sort of commonly accepted best practices for IT diagnosing. You never get these things 100 or even 80& perfect: look at all the contention in hardware benchmarking. On the other hand, having something is better than nothing at all.
While it's unfair to load more work on the desk of the newly formed Open Management Consortium, this topic is something that'd fit well with their goals. One of those goals, as I see/understand it, is establishing interop, including semantically, between different OSS systems management stacks (and comercial if The Big 4 want to play along).
As for Cisco, I must admit that I don't know them, their plans, and intentions well enough to even speculate. But, climbing the value stack into the C*O office is a safe bet for most any systems management vendor, open or closed source. As with so many pieces of software, once you get a hold of all the raw data, you want to do something valuable with it beyond respond to troubling states. Doing something with that data that involves the business side of the house — like BSM — is not only more exciting than alerting when a file system is full, but worth more money.
Most vendors want to have a verticle hold on the IT in their customers shop. Who doesn't want to be Microsoft?
In this modern world the art of Management has become a part and parcel of everyday life, be it at home, in the office or factory and in Government. In all organizations, where a group of human beings assemble for a common purpose irrespective of caste, creed, and religion, management principles come into play through the management of resources, finance and planning, priorities, policies and practice. Management is a systematic way of carrying out activities in any field of human effort.
Its task is to make people capable of joint performance, to make their weaknesses irrelevant, says the Management Guru Peter Drucker. It creates harmony in working together – equilibrium in thoughts and actions, goals and achievements, plans and performance, products and markets. It resolves situations of scarcity, be they in the physical, technical or human fields, through maximum utilization with the minimum available processes to achieve the goal. Lack of management causes disorder, confusion, wastage, delay, destruction and even depression. Managing men, money and materials in the best possible way, according to circumstances and environment, is the most important and essential factor for a successful management.