Government Operations Committee on Feb. 13th, 2012

[The Chair, NDP, Pat Martin]

Ladies and gentlemen, we will call the meeting to order. This is the 26th meeting of the Standing Committee on Government Operations and Estimates. Just before we move to the orders of the day, I'd like to recognize the Liberal vice-chair, the Honourable John McCallum.

[John McCallum, Liberal, Markham—Unionville, ON]

Thank you, Mr. Chair.

I'd like to introduce a notice of motion today for discussion at the next meeting, which states:

That the Standing Committee on Government Operations and Estimates conduct a study into allegations by the Defence Employees Union regarding contracting practices at Defence Construction Canada, and that the Committee report its findings to the House.

[The Chair, NDP, Pat Martin]

Very good.

Thank you, Mr. McCallum. The notice is served.

Therefore, we will move to our orders of the day. As part of our study on Shared Services Canada, we'd like to welcome the authors of the PricewaterhouseCoopers report that was, in fact, the genesis of the Shared Services Canada initiative. We welcome Mr. Ivan Milam, the director of PricewaterhouseCoopers, and Mr. John Chang, a partner.

Welcome to both of you, gentlemen. Thank you for being here. It's customary to give opening remarks, if you choose to do so, for five or 10 minutes, and then we will go to five-minute rounds of questioning about your landmark report.

[John Chang, Partner, PricewaterhouseCoopers]

Thank you very much, Mr. Chairman.

Thank you for having us. We're pleased to be here to answer any questions this committee may have.

By way of background for this report, the origin of this study was a Treasury Board Secretariat request to PWGSC to study the future of data centre services delivery within PWGSC, with an eye to extending the result of the study to the whole of government.

In January 2010 PwC was engaged by PWGSC and the Treasury Board to study the future of data centre service delivery. The scope of data centre services to be studied included facilities, servers, storage, and mainframes, but did not include networks or applications. The original scope of the study focused on PwC and included development of data centre requirements, future state delivery options, and options analysis, including case studies and recommendations for options and road map.

The scope of the study changed four months into the engagement and included all of government.

The study was undertaken in a very consultative manner and had inputs and participation from across departmental committees, from directors general of these departments, and from the CIOs of these departments. The study itself also had robust oversight. The cross-departmental committee of CIOs, which met monthly, provided project governance, and we also had two external expert advisers who reviewed all of our deliverables and provided reports directly to the crown.

That's the end of my opening remarks.

[The Chair, NDP, Pat Martin]

Thank you, Mr. Chang.

We will go, then, to our first round of questioning.

On behalf of the NDP, the official opposition, we have Alexandre Boulerice.

You have five minutes, Alexandre.

[Alexandre Boulerice, NDP, Rosemont—La Petite-Patrie, QC]

Good afternoon, gentlemen. Thank you for being here today. We're pleased that you could join us for our study on Shared Services Canada.

My first question is simple. How much did your study cost the government?

[John Chang]

Our fees to the government were just in excess of $2 million.

[The Chair, NDP, Pat Martin]

I'm sorry, Mr. Chang. Could you say that again?

[John Chang]

Sure. PwC fees for this consulting engagement were in excess of $2 million.

[The Chair, NDP, Pat Martin]

I see. Thank you.

[Alexandre Boulerice, NDP, Rosemont—La Petite-Patrie, QC]

If the savings created by centralizing data and services were $45 million, that would mean that your study would have cost 5% of future savings. I think that's huge.

As for potential future savings, you estimated that the cost reductions would be between 6% and 36% of the amounts currently committed, or $45 million to $293 million, which is a fairly large margin. Why be so cautious? Why such a margin? Is it because it's difficult for you to really see how much we'll save?

[John Chang]

There are two parts to our report that contribute to forecasted savings.

First is transformation. Within our report, transformation refers to consolidation and rationalization of data centre facilities and to server rationalization, consolidation, and virtualization, as well as virtualization and rationalization of both mainframe and storage towers, as we call them. That's one area of savings potential.

The other is transitional savings, or a transition to what we call alternate service delivery, either to a shared services centralized internal delivery organization or to a private sector third party provider.

In both of those cases, we did a bottom-up analysis to determine what the potential savings may be. Within the transformational area, for example, the server consolidation benefits were estimated based on different assumptions we took on the current capacity of those servers for consolidation.

For example, if you had servers running at 20% in one department, and for a very similar application in another department it's 30%, and you put it into the same hardware platform or server, you can see that there are savings there. That's on one end of the spectrum of savings. If we took a range of consolidation savings of, say, 50% or 60%, the range was used because we did not have the opportunity to actually do very detailed data gathering or look at capacity utilization of servers, for example. That's why the utilization of the range was used.

In the transitional state, we determined benchmarking prices for external private sector prices for these services, which had a range—as you know, there's not an absolute out there for that—as well as looking at best-practice internal current service delivery. For example, the mainframe area within the Government of Canada is very well managed, in our opinion, and its price points are very, very attractive, so we use that number as one of the benchmarks to see what the benefits would be if we moved to that kind of delivery model.

In both of those cases, we were able to put a range both on a transformation and on a transition, which resulted in these kinds of conservative and optimistic forecasts.

[Alexandre Boulerice, NDP, Rosemont—La Petite-Patrie, QC]

I'll be cautious in this case. I understand the challenges you were facing. Perhaps you didn't have all the data required. I know, however, that when I was in elementary school, my teacher wouldn't have been too happy if my answer was 293 million, but maybe 250 million less. That's quite the range.

You also said in your report that the private sector should be involved in Shared Services Canada at a rate of 54%. That's quite a big role for the private sector. How do you justify that?

[John Chang]

Again, our analysis had two aspects. There's a transformational part and a transitional part. In both cases, we believe that it's best practice to leverage the best expertise wherever it may be, whether in the private sector or in the public sector, to undertake those two transformational or transitional journeys.

With respect to the 54%, we actually looked at what we call towers or service bundles, and that's well documented in our report. They include things like mainframe operations or server operations, common infrastructure server operations, data centre facilities, and so forth. When we did the analysis of those requirements against the private sector capabilities, we found that the areas that the private sector can deliver well are being commoditized as a utility in the marketplace. When we did that analysis, we came to this 54% as being the amount of effort or the amount of work being done in this space that can be done by the private sector in the most efficient manner.

[The Chair, NDP, Pat Martin]

Alexandre, that concludes your time. Perhaps you can return to that area in future questioning.

For the Conservatives, we have Bernard Trottier.

You have five minutes, Bernard.

[Bernard Trottier, Conservative, Etobicoke—Lakeshore, ON]

Thank you, Mr. Chair.

Thank you, guests, for coming in today.

I want to talk further about the data centre consolidation. Would you describe data centre consolidation as the low-hanging fruit in this overall systems transformation and IT transformation that the Government of Canada is undertaking?

[John Chang]

Absolutely. It is an area that is quite federated, if I can put it that way. As we documented, we have over 300 data centres and well over a thousand of what we call “points of presence”, where servers that are delivering services or running applications for the Government of Canada are located. It is a logical conclusion to look at it. That kind of environment is very difficult and costly to manage and is easily identifiable to be consolidated from a facilities perspective.

[Bernard Trottier, Conservative, Etobicoke—Lakeshore, ON]

If it is fairly low-hanging fruit and fairly easy to get at, why wasn't it done sooner? What kind of organizational inertia would have prevented the government from doing it years ago?

[John Chang]

I think part of the reason, obviously, is in the structure itself, from what we observe.

One of the things I would also point out, though, is that when we talk about data centre services, facilities make up only about 6% of the total spent. Visually and intuitively it sounds like a big opportunity, but compared to the actual dollars that are being spent within data centre services, it's a relatively small expense. There's a trade-off there, for sure.

[Bernard Trottier, Conservative, Etobicoke—Lakeshore, ON]

What does data centre consolidation enable in terms of further savings in terms of applications, disaster recovery, data backup, and those kinds of things? Can you describe some of the follow-on benefits that happen once you consolidate data centres?

[John Chang]

Sure. When you're looking at data centre consolidation, it should not be looked at by itself. IT is a very integrated function, which includes, as you just mentioned, infrastructure services like servers, mainframe, storage, and all of that, as well as applications that run on it. If data centre consolidation is done properly, it will have a transformational impact on other areas. To get the full benefit of data centre consolidation, you must look at what's inside the data centre and you must look at what's inside those servers, which includes hardware, software, and applications.

[Bernard Trottier, Conservative, Etobicoke—Lakeshore, ON]

Would data centre consolidation be a necessary precondition for email consolidation or for standardizing certain applications that might be on different platforms, for example?

[John Chang]

From a program perspective, given that it is transformational in nature, you want to undertake initiatives in one area to fuel the success of the next stage, for sure. From a technical point of view, it's not necessarily that one is dependent on the other. You can in fact do an email consolidation that is independent of the floor space itself. From an overall program point of view, it's probably wise to look at this as a series of gates, as we described in our report, and make sure you're succeeding in one in order to fuel the next stage of the evolution. That means doing the data centre as one of the consolidation opportunities--as you call it, low-hanging fruit--which will then lead to the confidence and maturity for the Government of Canada to actually take on the next gate, as it were, which would be server consolidation, mainframe and storage consolidation, and ultimately--which is the real prize--potentially, application consolidation.

[Bernard Trottier, Conservative, Etobicoke—Lakeshore, ON]

Thank you.

How much time do I have, Mr. Chairman?

[The Chair, NDP, Pat Martin]

You have ninety seconds, Bernard.