IBM late last week quietly celebrated the 10th anniversary of its acquisition of Tivoli Systems, marking a 10-year roller coaster ride in the network and systems management market.
Although IBM in the early years made significant profit on its $743 million acquisition, it struggled with internal integration headaches along with even bigger customer integration headaches that caused deployments to stretch for as long as 24 months.
After the dust settled in the enterprise management framework wars between Tivoli and large competitors such as Hewlett Packards OpenView and CAs Unicenter, the word “framework” was dropped from everyones vocabulary.
And customers turned to simpler point tools to solve the most pressing problems. But, after being missing in action from any IBM growth trajectory, Tivoli has come back in the past couple of years to post solid double-digit growth (15 percent last year), take the drivers seat in IBMs autonomic computing initiative and execute on a series of acquisitions, including the most recent acquisition of MicroMuse and its NetCool network management software.
Senior Editor Paula Musich talked on April 4 with Tivoli General Manager Al Zollar about the ups and downs, and the challenges ahead, for IBMs management software unit.
Tivoli and IBM together have had a bit of a checkered past. At one time it was the darling of IBMs software portfolio, and at other times it has been a big black eye for the company. What lessons did IBM learn from the lean Tivoli years?
The biggest lesson we learned is that quality matters in enterprise software. The quality of the functionality [then] was not what it needed to be. The technology was in transition. When we had a team pinned down fixing quality problems it was hard for them to transition to Web [technologies] and J2EE [Java 2 Platform, Enterprise Edition] and now SOA [service-oriented architecture] architectures. It made it hard for the team to evolve to those things while they were focused on fixing quality and consumability. If you miss a beat it can hurt you for a number of years.
After the bruising that all the framework providers took, everyone stopped using the word framework or platform. But how much of that technology is still used in the products that IBM sells today?
They are used quite extensively. For those customers who made early investments in our products, our job was to help them get value out of them. We continue to support those customers. Of course, were now presenting them with alternatives that represent upgrade paths to high quality. Tivoli Monitoring Version 6, which started shipping in the fourth quarter of last year, offers a nice automated upgrade path for customers who used the distributed framework monitoring product.
That gives them an end-to-end monitoring product [that spans from the mainframe] zSeries to Windows, Linux and Unix with a consistent set of visualization and policies. It is a journey our customers will be on and we need to make each step an easy and logical step to take.
It used to be said that for every dollar spent on a license, another $3 or $4 was spent on services. Is that ratio still accurate today?
It depends on what industry reports youre reading. Some will go higher. Enterprise software in general, when youre dealing with the issue of complexity of composite applications, most large enterprise customers have a minimum of 500 Web tier applications, with some as many as 2000. Integrating any software into that environment requires know-how. It is not shrink wrap territory. For the know-how associated with bringing to life the capabilities of Tivoli products, we work hand-in-glove with [IBMs] Global Services and industry partners.
What Does the Future
Hold for Tivoli?”>
How would you characterize how Tivoli is viewed today by IBM upper management and by customers and the market?
I think weve certainly been viewed as a key brand within our software business by IBM and probably more significantly by our customers in coping with complexity. Technologies like virtualization, servers, storage and networks—we help them with that, but the logical complexity remains. That complexity is a big issue.
Compliance is a huge issue we hear. The rate of change is a big issue too. We see customers seeking better solutions to manage IT deployments in much more business-oriented fashion, rather than an ad hoc fashion. You have to be linked to the overall business goals. Thats what our IT Service Management strategy is about.
Speaking of your ITSM strategy, some detractors believe Tivoli is at a disadvantage because it has no service desk offering—despite the fact that it is a key part of ITIL. What exactly is Tivolis service desk strategy and how can Tivoli communicate that clearly to customers?
Our strategy in a word is integration. We think service desk is a component of IT Service Management, but you cant tell customers to take the one they have out. Our strategy is simply integration. Our change and configuration management database [CCMDB] is an open structure that lets us federate information from [multiple sources]. Anyone who creates a service desk knows it has to make an interface available to [different] data structures. Thats been very effective for us. Weve had two years of double digit growth. Thats a proof point to the effectiveness of our strategy.
How is work progressing on the Tivoli CMDB effort?
Weve been working on this before we launched our IT Service Management strategy in May of 2005. We had a number of early customer design partners like Ford that have been giving us feedback and we are moving software into deployment. General availability is in June and everything is on track.
CMDBs have become a huge topic. Everyone in some part of the management market has one now. How do you differentiate as everyone talks about CMDB?
Everyone is requiring that their database is it. But the CMDB really is a federated source of configuration items about end-to-end infrastructure—not data specific to one management application. With point products there are conflicting data on things like servers, storage devices, networks, incidents. These configuration items are all over the place. One database specific to asset management as the one CMDB is false. It has to be based on open standards and mask the management of master data. There are other differentiators: we package core processes. Change management process and configuration management process is packaged with our products. Theres no proprietary workflow engine.
What is the most significant effort underway at Tivoli now?
We recently introduced a full line of [Tivoli] Express products aimed at the midmarket. We worked for a year and a half to really build the packaging and pricing needed to be more consumable through the channel by midmarket customers. Thats going to be a substantial improvement to address the market and gain market-share.
Also theres the acquisition of MicroMuse. We found that the NetCool team at MicroMuse are the worlds deepest specialists in managing the most demanding networking environments on the planet. They are present in all major telcos, high-end financial services organizations and within [the Department of] Defense in war fighting functions where mission critical takes on a whole new meaning.
Theyve given us a measure of technology and skill in the most demanding networks well integrate into the Tivoli portfolio.
Whats the roadmap for integration of MicroMuse into Tivoli?
Weve talked to clients individually to describe it and get their input. When we announced the MicroMuse acquisition, we said within the first few months we will have incredible levels of integration at a visual level and a data level. Because of our collation acquisition, the ability to bring data in is really already there. Weve also integrated Netcool Omnibus into the Tivoli Enterprise Portal, so data integration and visualization is there. The key point to make is this: were approaching this the same way we did with [IBMs acquisition of] Candle. Every step should be a release upgrade, like upgrading to a new release of software that just happens to have a blend of software we and MicroMuse built.