I spent a few days at IBM Pulse last week, the IBM conference associated with infrastructure management and the Tivoli product line. It was an enormous event at the MGM Grand in Las Vegas – 8,000 people as announced at the conference.
IBM is of course in a class by itself in terms of its breadth of vision. This was not a conference on IT management; it was a glimpse of the multidimensional technological future hurtling towards us. Health care, smart cities, utilities – one was as likely to meet an expert on muncipal power distribution as to meet an IT service desk manager. The keynotes by IBM executives hit a set of IBM themes that are becoming familiar: Analytics, Watson, health care, DevOps, Smarter Planet, and IBM’s formidable hardware and software. They are developing quite a view of a thoroughly IT-enabled society. But I’ll leave macro-analysis of IBM to others. My primary questions in attending Pulse were:
1) Is IBM finally integrating its formidable portfolio of IT management products into a coherent IT value stream capability?
2) What’s Maximo?
Towards these ends I found myself favoring the show floor over the breakout sessions. I spent quite a few hours working the booths, meeting a number of fascinating vendors and listening in as customers posed various problems for IBM and its partners.
Of the major IT vendors, it’s been hard of late to spot IBM’s presence in core IT service management. This is strange, as IBM has some of the industry’s deepest roots there. The IBM “yellow books” (A Management System for the Information Business) were an important precursor to ITIL in the 1980s. However, after the IBM sale of the Tivoli Service Desk to Peregrine in 2000, IBM had essentially nothing in the Service Desk space for six years, until it acquired MRO Software, the maker of Maximo, a major vendor of Enterprise Asset Management software.
The Maximo product filled an essential gap in IBM’s portfolio, being a flexible workflow framework capable of serving as a general development platform. A core set of ITSM products have been developed using it, yet Maximo still seems to have more traction in its original niche of Enterprise Asset Management.
Now, Enterprise Asset Management (EAM) is an interesting field, covering a wide variety of capital-intensive problems from refineries to rail transport. Like its sister IT Asset Management, it is NOT the same as the Fixed Asset modules one sees in a Financial ERP system; in addition to tracking detailed asset financials, EAM addresses issues like maintenance, uptime, regulatory compliance, and so forth.
To provide further perspective on Maximo’s reach in EAM, the following user groups met during Pulse:
- Maximo Life Sciences User Group
- Maximo Airports User Group
- Maximo ITSM User Group
- Maximo Manufacturing User Group
- Maximo Transportation User Group
- Maximo Facilities Management User Group
- Maximo Oil and Gas User Group
- Maximo Utility Working Group (MUWG)
Not your traditional IT Service Management suite by a long shot! This breadth has been further bolstered by IBM’s acquisition of facilities & real estate solutions provider Tririga. The overall message to the market is clearly, “IT is becoming pervasive, mobile, and embedded, and what’s important is how it’s applied.”
One of Maximo’s significant aspects is the depth of its partner ecosystem. Represented at Pulse was a rich array of vendors skilled in implementing solutions on the core Maximo framework. This translates to a deep pool of Maximo expertise that can be leveraged across domains. I talked with at least one vendor who had originated in an unrelated industry vertical and was now building ITSM-related solutions on top of Maximo. Some of the conversations I had:
- I know from my PeopleSoft years that very often the SDLC tools provided with Maximo-like frameworks are rudimentary at best. Both interloc solutions with their Interloc Release Manager and Asset Partners with their Config Compare are addressing this gap.
- gen-E offers Maximo extensions for runbook automation, social media, knowledge management, and other areas. Of interest was their process analytics capability, reflecting some things I’ve discussed here.
- Regular readers know that demand and execution management for IT are ongoing areas of interest for me. I was therefore particularly intrigued with the Maximo vendors building products in this area. In particular, Pipeline Software with their Syntempo and Primavera integrations was notable, although not targeted to IT management. CiM’s Visual IT Scheduler was the kind of innovative solution that I love to see emerging from a framework ecosystem. I am a big fan of this kind of visual view on IT demand and execution; I’ve seen it have C-suite traction. Had a good discussion with CiM’s Sebastien Charbonneau as well about the potentials of true production scheduling for IT work.
I still am amazed that more operations perspective has not transferred to IT management. When I read (for example) marketing literature from Pipeline stating “Pipeline Software’s Syntempo® solution automates the craft feedback loop, eliminating the need to chase down work status” I ask myself, how is such a capability different from managing a team of systems engineers in a data center? Other than the term “craft” which (I infer) refers to the kind of plant maintenance personnel who work from big Snap-On carts instead of at terminals?
IBM realizes this, and the “IT/OT” (Information Technology/Operations Technology) convergence was mentioned a few times from the podium and in one on ones.
Of course, the major Maximo news for IT service management is the announcement of the SmartCloud Control Desk, a consolidation of Tivoli Asset Management for IT (TAMIT), Tivoli Service Request Management (TSRM), and Tivoli Change and Configuration Management Database (CCMDB). This has been accompanied by extensive usability engineering which is good; usability consistently comes up in EMA discussions as a challenge for Maximo-based ITSM products. Watch for further in-depth discussion of SmartCloud Control Desk by EMA.
On the other topic: IBM has long had waterfront coverage of the IT lifecycle from development through operations, but this has seemed less than the sum of its parts. Along these lines, the concept of DevOps was widely discussed, including at a special analyst presentation, and I spent quite a bit of time at the back of the Expo floor with Rational and Tivoli engineers familiar with the details of IBM’s support for this movement.
IBM has turned to the Open Services for Lifecycle Collaboration (OSLC) standard for some support in this area, but there are details that this spec does not address. I observed a Rational engineer developing a simple topology representing an application server talking to a database, and I asked “what metamodel is that”? He said it was (unfortunately) not a standard, but a proprietary construct in the Rational Software Architect for Deployment Planning that everyone referred to as “Zephyr.” (It IS based on Essential MOF, for those who care, so it interoperates with the OMG stack.) He also told me that from his perspective, there were still some unresolved complexities around the relationship between model, code, and deployment and their supporting systems (UML, source code management, and CMDB/discovery). I didn’t have time to drill into the last mile of what he was saying, but it seems they are close to closing the gaps.
I’ll take some credit here. A Google search for Model-Driven Configuration Management yields nearly 4 million hits, with my eight-year-old post proposing the concept at the top. There are any number of reasons why such theories have struggled, from Agile to dynamic virtualized environments, but I still believe that there is an important and as yet unrealized service portfolio information flow between development and operations, and I continue to look at specific DevOps practices for a solution.
Random: thrilled to be interviewed by Jan and Grady Booch (that’s him and me in the picture) for their new series, Computing: The Human Experience. (I talked about my first experiences with computing, courtesy of the historic Minnesota Educational Computing Consortium and related initiatives. I played Oregon Trail version 1. And was a big fan of the chat facility on the Control Data computers known as “XTalk.” Been doing social computing since 1975…)
Also pleased to see my book in good quantity at the Pulse Bookstore. Forrester analyst Glenn O’Donnell, a classy guy if there ever was one, bought a copy and asked for my autograph. (I own his book on CMDB, so I guess fair is fair.)
Finally, cave canem: Large companies like IBM are able to be their own test beds for their IT management solutions, a strategy known as “eating one’s own dog food.” IBM SVP Steve Mills used this metaphor during his keynote, and also came up with the astounding visual image, “zSeries Linux runs like a scalded dog.” Yelp!
One hopes no dogs were harmed in the making of these metaphors. I’m not sure I’d want to be Steve’s pooch, between having my food eaten and getting hot water dumped on me. But I won’t be the one to report him to PETA. Although I can just envision Pamela Anderson showing up with a protest sign at Pulse.
For additional IBM perspective from my EMA colleague John Myers, see “Focus On The Forest Rather Than The Trees:IBM Smarter Cities China.”