Marc Plamondon has worked for the Province of Quebec for over 25 years, and has held his current post of mainframe software manager for the past two. He has no background in IT, but Plamondon's leadership skills have been recognized with the 2007 Share Award for Excellence in Technology.
Plamondon led a team from the ICT Directorate-General in Quebec's Centre for Shared Services on an ambitious undertaking to migrate 200 Oracle databases from Unix to Linux on IBM's System z. In conversation with senior writer Lisa Williams, Plamondon talks about why the project was considered so political and how he managed to win hearts and minds. He also reveals his "techie" side.
Q. What was really the key objective of this project; why did you make the switch from Unix to Linux?
A. We looked at the total cost of operation (TCO) of three platforms - Unix, Linux and Windows - and we discovered that with Linux we were saving roughly 30 per cent compared to the other platforms. That was one of the main reasons.
But the key reason was improving agility. The TCO is secondary to us now. When we first started the project, the savings were more important. But now as we're down the road in the project, the agility is more important for the higher management than the TCO. They don't talk to us anymore about saving money because we've proved that point to them. Now we're more interested in the agility on providing solutions to the customers.
Q. What kind of operating problems were you experiencing before you made the change?
A. We had many issues to solve. There were a lot of operating systems on the Unix side, so there were many systems to evolve. We had unsupported software and were not as satisfied with the disaster recovery. There was also a fast-growing environment - roughly 100 servers per year of growth - for Windows and Unix as well, and that's been the case for at least a few years.
We were under-staffed to maintain these "solutions" a few years ago, so we had to undertake a big consulting contract to help us support Unix. As you know, it's more expensive to have consulting personnel than internal personnel, so in the long run, with 30 per cent savings on the Linux side, it's going to help on the TCO. And we needed to have flexibility for rapid deployment of customer solutions.
Q. When did you start the migration and what stage are you at now?
A. We began the project in 2003 with a proof of concept, from end to end. We had a z800 mainframe made available and then we installed the software to fulfil that proof of concept. We also installed an application - an SAC (software assigned controller), which is a Net-run, reusable software solution - and we had to test that. It was good in terms of the functionality and stability, and the transportability was also tested.
Q. If operational cost savings kickstarted this project, can you be more specific about the benefits that began to unfold, in terms of "agility" and improved performance?
A. We definitely saved money on the business case: within three years we're going to have repaid the new machine (it's a z9 Enterprise Class).
We reviewed our business case recently with other technical personnel, and we'll do better than anticipated because we have added more databases than were planned. I think that is the main reason we added many more data-bases. We didn't first install them on the Unix environment and then migrate them. We installed them on the z/VM (z Series virtual machine) environment right away, and that improved the business case results.
The performance was not much of an issue for us; it was more the agility and the TCO, because we were quite confident on the performance side. Also, we have very good performance on the z/VM Linux environment. The database we run is for the government portal, so if it didn't perform well, we'd know about it really quickly.
Q. With respect to the award you won, what does that kind of recognition mean to you and how does it rank against other awards you've received?
A. We're very proud of it, and that project was really like a cultural shift in the way we work. With the award, people are more accepting of what the project represented.
It gives more credibility to the project in-house, with the consulting personnel and even the internal personnel who are mainly working on Windows or Unix. It's really a matter of pride for the people who identify themselves with the technical solutions. And on the outside it gave the Province of Quebec visibility and credit for what we did.
The Directorate-General of Information and Communication Technologies (DGTIC) also won an OCTAS award (granted by the FC)dC)ration Informatique du QuC)bec) for our change of address solution about two years ago. But when I worked on that project, it was only on the technical side, and it was mainly the user side that was recognized.
The Share award recognizes much more than only the technical side. It's also a much broader recognition: the OCTAS award was mainly provincial, and this one is international - it's more important.
Continued: A project of culture and change management
Related content:
Spotlight Central
Finetuning shared services
Something for everyone
icon.
