We demand agility from our IT environments and offering software solutions to address the issue. How about infrastructure? Understaffed and bogged down by daily fire-fighting, can we truly expect the infrastructure team to complete a project without an effect on other groups and at the speed expected?
Contributors

• • Wallace Dalrymple, Chief Network & Telecom Architect, Global Technology Management, General Motors
  • Johna Johnson, President, Nemertes Research

Transcript

Sanjog Aul [00:00:00]:
Good afternoon and welcome to CIO Talk Radio. To learn more about the show, please visit www.ciotalkradio.com. Today’s topic is Agile Infrastructure: Is It An Oxymoron? Our guests for today’s show are Wallace Dalrymple and Johna Johnson. Wallace is the Chief Network and Telecom Architect, Global Technology Management with General Motors and Johna is the President of Nemertis Research. Good afternoon panel, welcome to this show.

Wallace Dalrymple [00:00:29]:
Thank you.

Sanjog Aul [00:00:32]:
So how were the last three quarters? I can’t believe that we already reached the fourth quarter.

Wallace Dalrymple [00:00:37]:
The year has definitely flown by really quickly.

Sanjog Aul [00:00:42]:
Now coming to this topic for today’s show, the fact that we already have a lot of problems and you happen to talk to folks who are in the infrastructure division, they feel that they have a tankless job, they work of endless number of hours, never get a break, but on top of it now there’s a renewed expectation of them making this infrastructure agile. That is to literally react overnight on based on a business need, maybe because of the software being changed and or anything else that needs to be done in terms of expansion, et cetera. So given the fact why in the first place we would even want to go that route, let’s challenge the fact that we could just plan ahead and Kale versus trying to say hey, we’ve got something new coming up, let’s totally switch gears and then even before that let’s talk about what does the world mean when they talk about the word agile infrastructure. What are the typical example.

Wallace Dalrymple [00:01:36]:
Sanjog, the infrastructure today enabled to support an agile infrastructure that really aligns with your business and the ability to change with your business is now critical more than ever. It is really an integral part of making those transactions or those business processes happen. So there are definitely an alignment, if there wasn’t in the past, if it’s not stronger today, of actually the business processes and the agile infrastructure supporting those business processes. So when you talk about scalability and being able to plan ahead in today’s world, it’s not always the capability to be able to plan ahead, but there is the expectation that your infrastructure can fail to meet whatever those business demands are.

Johna Johnson [00:02:30]:
Just to circle back to that question, I certainly agree, Wallace, with what you said and circling back to your question, Sanjog, when I think of agile infrastructure, I think of in the networking space things like convergence so that instead of having to pull separate voice and data circuits, you’ve got one type provision, a new user, you simply click the mouse button instead of configuring manually configuring something and in the storage and server space issue where Grid computing, Blake computing, Virtualization and Virtualized storage are examples of agile infrastructure and again that sort of salient characteristic, it being able to configure very rapidly to address dynamic demand and that’s how I think of agile infrastructure. Well, is that kind of what you mean when you think about it?

Wallace Dalrymple [00:03:13]:
Definitely. There’s almost a decoupling from the business processes and the application infrastructure that it is an infrastructure to be there today, so it has to be able to scale on demand and the technologies that you mentioned, grid computing and virtualization, those are the building blocks to enable that agile infrastructure to move forward.

Johna Johnson [00:03:37]:
And I really like that word that you use, decoupling, because what I think about the link pen of any kind of virtualized infrastructure or agile infrastructure, the fact that you separated out a reliance on a particular technology or particular piece of physical gear, so now you’re not necessarily tying voice services to a particular voice circuit anymore. So that’s a really good way to think about it.

Sanjog Aul [00:04:02]:
Now let us quickly look at the fact that we spoke about scaling on demand. How about extending on demand because any environment needs to scale as well as extend on demand and that’s what in true sense an agile. That’s how you would define agile in some cases. So would that be true in infrastructure also? Or would you say we have to still confine our definition of being agile in terms of infrastructure as just scaling?

Johna Johnson [00:04:26]:
I think being able to grow with demand, it’s actually a very key component and that dynamic growth. In fact, that makes me think of two things. One is an example that I always use of not agile infrastructure, which is a client of ours whose name shall be obscured for obvious reasons. That was running into a real problem which was that they were too successful. This was a large retail organization. They were opening up or attempting to open up a couple of sites a month and what they were finding was that because they were using extremely gated technologies for voice and data, they were having a hard time getting the circuits in place and having a very hard time getting everything set up and configured.

Johna Johnson [00:05:03]:
So IT was actually holding back the business’s ability to grow and to me that’s an example of non agile infrastructure and the real challenge to not if you’re choosing not to adopt an agile infrastructure, is you’re basically setting IT up as being the gating factor and you’re predicating your business growth or your business adaptability on your IT capabilities, which is never a very good place or comfortable place for an IT person to be.

Wallace Dalrymple [00:05:30]:
Yes and if I could build on that, I agree you don’t want the goal is not to have your IT operations department or your infrastructure to be the roadblock to adopt new business processes or business change. An example that I can think of is if you’re an e-commerce, maybe a global e-commerce organization or you’re a financial institution. The business and even the customer centric driven ability to provide services around the clock no matter what country you’re in or what time of the day it is really key in order for enterprisers to be successful. Today our customers expect, for instance, online banking. I expect that when I get on, no matter what time of the day, that I’ll be able to access my account information and if for some reason, and maybe some of us have experienced a message that says sorry, we’re down for maintenance, when I view that message, I don’t see that as being an agile infrastructure because while I might want to look at my account information sunday night at 10pm, I still expect from a customer perspective to be able to access that information.

Johna Johnson [00:06:47]:
Absolutely and I think that real time responsiveness is one of those characteristics about agile applications and agile infrastructure. I would show one second that when you look at the business value, it’s the ability to respond very rapidly to changing business requirements. Whether it be I suddenly got a sudden surge of customers at 4 o’ clock on a friday or oh my god, my boss just called me, we’re going into Indonesia and I need to have everything up and running in two weeks, both of those examples of really needing to respond quickly to real time changes in the environment, I think that’s probably the defining characteristic of an agile infrastructure.

Sanjog Aul [00:07:23]:
So when we talk about having a stable environment and also we want the growth, how do we take the two together? As we always say, like a software, there’s a bug. So you report a bug, fine, they will fix it. They will take you through different stages of environments like testing, staging, et cetera and you’re done. In infrastructure, yes, of course you could have the same stages, but in the first place they don’t want anything to be down and you only get that one hour window to get your major rollout. So are we not putting on unrealistic expectations on infrastructure?

Wallace Dalrymple [00:07:54]:
I don’t, in today’s market that’s not unrealistic in my opinion that is more of a requirement. For example, if you have an outage or you’re affected by any type of infrastructure, whether it’s related to internal or external entities, you have to have the processes and the technology in place to be able to handle that incident, because it could eventually or could result in lost revenue to your business and in today’s competitive global economy, that could be detrimental.

Johna Johnson [00:08:27]:
And Wallace, I know you like to talk about the people and process piece as well as the technology, and I figured that was what you were going to respond to, and in fact, I think that’s very key, is to really understand what agile infrastructure requires from people and processes, not just from the technologies. Just a second, Wallace, what you said there is, when we benchmarked the disaster recovery requirements of a number of very large financial services firms, we found that over half of them said that the acceptable amount of downtime for their organizations was none, zero deaths, they weren’t allowed to go down periods, full dot, and they had to architect their infrastructure in such a way that it didn’t ever consistently go down. We called it, we like to joke about this as being really extreme availability, but that the reality is, he said, “oh, it’s been dead”.

Wallace Dalrymple [00:09:16]:
Exactly and one of the continuing challenges is once you build that infrastructure, I think the challenge doesn’t go away. It’s the thing in that infrastructure because as your business changes and gets more dynamic, your infrastructure has got to be able to match that page and that is a challenge I think that every organization faces.

Johna Johnson [00:09:35]:
And yours in particular, because by heaven, I drive a GM car, and if I want to get a new one, I want to make sure that you guys are working 24, seven day and night making sure that the infrastructure is in place to get me one.

Wallace Dalrymple [00:09:47]:
Exactly and we are working day and night.

Sanjog Aul [00:09:50]:
Talking about the fact that we have an expectation that we simply cannot not have a zero. We should have a zero tolerance policy in terms of the downtime, in terms of infrastructure for such a global competitive market. What is that doing to the actual workforce, which is doing the job? Are they able to cope up with it because I’m trying to bring in the human element, while on one side it’s processes and people who are supposed to be good but how are we managing that in most cases because most companies or most listeners who are out there, they may not have that thing happening with them. They might be having their infrastructure breaking every two days. So what are they doing wrong? What are the gotchas here?

Johna Johnson [00:10:35]:
This is a, that’s a tough one because without knowing the particular organization, it’s impossible to say specifically what they’re necessarily doing wrong but I think a good place, for starters, as well as where you let in with people, process and technology and just starting with the technology. In a lot of cases, one of the problems with people are wedded to legacy technologies that typically don’t have that degree of flexibility. So for an example, if you are not yet moving to virtualized server environments using VMware or similar products from other vendors, your server provisioning time is probably too long. Similarly, if you’re not moving towards a converged voice and data infrastructure, your ability to provision voice and data communication services is probably limited by your technology. So the first place I’d start looking, if somebody said I’m having a real problem maintaining my environment at the level of availability, I’d look at which technologies are you using? Are you using outdated technologies? The next thing I’d look at is have you standardized effectively on the technologies in terms of product revisions, vendors,

Johna Johnson [00:11:41]:
have you really integrated this or do you have a mixed mismatched hodgepodge of lots of different technologies that you’re trying to glue together with baling wire and gum? They can do the kind of environment where typically you have pretty bad problems with availability and while is this making sense to you? Not that you guys would necessarily have these problems.

Wallace Dalrymple [00:12:00]:
No, we have no legacy applications in General Motor.

Johna Johnson [00:12:05]:
And no Baling wire, no glue gotcha.

Wallace Dalrymple [00:12:07]:
Duct tape,

Wallace Dalrymple [00:12:08]:
we don’t, we ran out of duct tape so we stopped using it. No accurately, I couldn’t agree with you more. There are definitely the legacy challenges out there and the technology component in working on the processes is a lot of organizations, including General Motors, have looked at standardized global common processes like Idle and trying to bring that framework, while it’s not an end all solution and isn’t advertised to me, one, it does provide some structure and some standard processes that organization can build upon and then they can customize those processes to their business model. So I think it’s important that when an organization or an IT infrastructure is experiencing lots of issues, there are technology solutions, or some way you can call them band aids, that will help how you handle those issues and how you standardize on the processes to deal with those to mitigate those downtime and hopefully if you build on the processes and the technology that gives you direction for the future to improve your customer service and your business.

Johna Johnson [00:13:15]:
Wallace, you mentioned Idle in particular. Is that a specific endorsement of the Idle framework or are there other frameworks that kind of serve in the same way and the reason I ask is we’ve benchmarked a number of enterprises and asked them which framework they’re using and in a lot of cases Idle is bubbling to the task partly because of the complicity.

Wallace Dalrymple [00:13:36]:
We’re using the standard Idle processes and we’ve trained all of our IT resources and employees on those processes, not only internally but also externally to suppliers and partners, so that we’re really seeking the same language. So when we talk about some of those Idle processes, change control and configuration management, it’s really we all understand what we mean when we mention those.

Sanjog Aul [00:14:01]:
We just mentioned about. I think Johna, you mentioned about legacy components in the organization and so that’s a reality. We and there is in fact a mandate sometimes that you have to preserve existing IT investments. What is that tipping point and which point you say, okay, we’ll let go of this because it’s taking way too much time or energy. There are political issues, then there’s of course financial. Like you can put numbers to say this is how much I’m putting on supporting it versus this is what I would need to do without us or buy a new place.

Johna Johnson [00:14:31]:
I think some respect the case by case decision of course you can’t just walk away from your legacy application but I think it’s important to really understand whether you’ve got a commitment to a technology, to a product or to a vendor, because those are different strategies. For example, if you have made a commitment in the application space to a vendor such as Oracle or SAP, but you want to get involved in moving to a service oriented architecture, you can put press bar on your vendor and say, listen, please deliver your five year roadmap for migrating to a service oriented architecture so that I can see if that aligns with my five year roadmap and make sure that we can move forward in an appropriate pace and if not, that’s going to jeopardize your standard standing as a favored vendor. Similarly, of course in the network convergence space, you may be looking at a more traditional player such as in the voice base, a Nortel or a Lucent or an Avaya, and in the data space, maybe a Cisco and you may be asking them show me how I can migrate from your solution to void over IT. That’s what you do when you’ve got a commitment to a legacy vendor.

Johna Johnson [00:15:35]:
If you’ve got a commitment to a legacy technology, I would actually turn around and say you want to revisit the cost of that commitment. For example, if you’ve committed to TDM void simply because the near term dollars and cents, you may want to look at that and say what the value in again dollars and cents to having an agile business enterprise and does it make sense to take a short term hit on say the amortization costs of your PBX in order to start seeing the competitive benefits of agility? Where you run into real problems is if you’ve got a commitment to a particular product because there’s a not a lot of leverage at that point. You can’t turn, you can’t turn stone into jewelry unless it already is jewelry and if you’ve made that level of commitment, you’re not really going to be able to. There’s not much you can do. I think the first thing to do is to really ask yourself what’s your commitment truly is to a legacy vendor, to a legacy product, the legacy technology and then your strategy in that light.

Wallace Dalrymple [00:16:37]:
I couldn’t agree with you more on that. You mentioned about legacy systems and I just recently read an article and about mainframes and COBOL, a language reprogramming language that we might have grown up on, but not one that is talked about too much today anymore but the amount of COBOL that’s in the industry and in mainframe today is more than there are people that still have the knowledge to actually use that programming code. So the forecast actually is that we will run out of COBOL programmers before we can actually get off of those legacy systems. Well, that’s a challenge I think a lot of IT organizations are facing today as well.

Johna Johnson [00:17:18]:
And economics being what it is, if you’re one of the COBOL programmers listening, you’ve got to be cracking your knuckles with glee because as the demand ratchets up, so does your price and that’s always a great place to take economically. Supply is going down, demand’s going up, you’re happy.

Wallace Dalrymple [00:17:35]:
Yes, exactly and also like a touchdown, you mentioned service oriented architecture and I think that is a component today, a technology component that businesses as well as IT is bringing to the forefront as a driver for the agile infrastructure because when I think of service oriented architecture and providing business component services that can be leveraged and can be expanded upon, you look at things like it provides standardization, automation, virtualization. So I think the service oriented architecture, while a buzzword and a lot of organizations are looking to see how they can benefit that new architecture, I think it’s pretty key in enabling the agility of the infrastructure.

Johna Johnson [00:18:23]:
I would completely agree and in fact I think it’s one of the major drivers because you can. What’s with the old biblical saying about not putting new one into old density you don’t want to put SOA onto an, into an old inflexible architecture because you’ll have all sorts of problems and you’ll lose a lot of the benefits that you would get from SOA but coming back to your comment about COBOL, actually there was, I was at a really wonderful session at one point where Cisco’s data center people were talking about the challenges they were facing in next generation data center architecture and one of the guys who was attending was a financial services person and he raised his hand and he said, how come Cisco doesn’t have mainframes in the data center and the data center guys just looked at each other and kind of scrubbed. Like why would we and what was very interesting is this is a very young company, relatively speaking, even though it’s one of the giants in our industry and in fact it outdated the era of the mainframe

Johna Johnson [00:19:17]:
and the reason I’m bringing that up is because one of the challenges with legacy infrastructure in general is your competitors will have made a commitment to newer, faster, more adaptive technology and for example, I remember when Cisco started to be able to close its quarter within something like 24 hours. A company can close its books from no time instantaneously. Part of that is that they didn’t have the hobble of legacy systems and legacy apps to tie into it. So coming back to what do you do with legacy systems? There’s a point at which you’ve got to say, listen, if there’s data here that we have to hang on to for either legal or compliance reasons, we’ll do that but in terms of the application itself, we are going to have to bite the bullet and move towards next generation technologies because even if we’re not, our competitors are and they’re going to be faster and more agile than we are and that’s going to put us at a disadvantage. So the cost of maintaining the legacy is not just the cost of those increasingly scared COBOL programmers. So for the cost of the competitive threat of people who are basing entire businesses on the next generation of technology.

Sanjog Aul [00:20:23]:
When we talk about infrastructure outsourcing, that’s another. It has been around for a while and that has been used as one of the solutions. If you were to take your become, make your company agile because instead of you trying to make, basically it’s a change management issue within the same company but if you’re outsourcing it, especially the areas which are out of control or you don’t want to control, then you just give them out and then try to integrate with them closely enough that you can potentially make your organization agile. How realistic is that option and are we in a way challenging that we don’t have an in house capability or the zeal to make that happen?

Wallace Dalrymple [00:21:02]:
I don’t not an in house capability issue or lack of skills or research issue is General Motors is actually working on their third generation outsource model and is really pioneering the outsourced model and making it effective for the business of building cars and trucks. It’s been extremely effective and it truly been something that General Motors has almost perfected and found the real advantages to having an outdoor enabling an outsourced model to haven’t impeded our ability or inflex or provided any inflexibility to provide business value today. So I would say that it’s not a lack of internal people skills, it’s more of leveraging your business model and making it work for your business goal.

Johna Johnson [00:21:56]:
I think that’s true. Although I think the one gotcha and I I’m sure all of you are probably not going to disagree on this one. The big gotcha is the siren song of outsourcing is I’m not gonna have to worry about an agile infrastructure. I’ll just give all my infrastructure over to somebody else and it’ll be their job to make it agile. There’s plenty of philosophy. One is it may be their job but it’s not in their best interest because at what point do they suddenly decide they being this sort of mythical outsource everything entity, at what point do they get the business advantage, the business advantage that you’re getting from the agile enterprise? In other words, if you’re saying my infrastructure is agile and I’m doing this in order to make my business agile, guess what? Your outsourcer actually doesn’t care if your business is agile. They care that you write your checks on time,

Johna Johnson [00:22:41]:
that’s it. One challenge is your interests are immediately divergent from your sources and the second one is part of the ability with carefully defined it early on. The definition of agility is the ability to quickly respond to changes. One of the things that makes it extremely difficult to match the outsourcing model and agility is what happens if you decide to change providers. What does the exit strategy look like and I’ve never met an outsourcer on the planet who will make it easy for you to leave. It’s usually slow and painful. The anticipate of rapid response to changing different conditions.

Johna Johnson [00:23:17]:
I’m not saying that outsourcing never works. Obviously it can work very effectively but I think that the trick is to understanding that these are some of the challenges and concerns and to really be very, I don’t want to say surgical, very precise about where you’re going to outsource and what you’re going to outsource.

Wallace Dalrymple [00:23:34]:
No, I agree with you. The way that we’ve handled it is within the IFNS organization, within General Motor and what we do with our partners, our suppliers, our vendors, who we outsource our IT services to is every IFNS VM employee has accountability, provides strategy and direction to those outsourced suppliers but we are sitting next to them, we’re working with them day and night. We spend a lot of time providing the information that they need so that they’re not waiting on us to provide direction. The communication is very tight and it really comes down to a lot of the processors that we have in place in order to make that work. Like I said earlier, we’ve gone through the third generation of our outsourced model which we’re implementing today. So we’ve got that down to where our suppliers are, our partners and whether they’re in the next building in another country or sitting right next to me. We really defined that communication.

Johna Johnson [00:24:39]:
What’s been the most effective way for you to keep their interests aligned with yours? For example, if General Motors meets its metrics for agility, do they get rewarded? Do they get are those metrics shared with them?

Sanjog Aul [00:24:53]:
Before you answer that question and Johna, before we finish that question, let us take a quick break. We’ll be right back after these messages. So listeners, please stay tuned. This is going to be an interesting conversations.

Sanjog Aul [00:27:24]:
Welcome back to the show, folks. For listeners who just tuned in, today’s topic is Agile Infrastructure Is It An Oxymoron? Our guest for today’s show are Wallace Dalrymple and Johna Johnson. Wallace is the Chief Network and Telecom Architect, Global Technology Management with General Motors and Johna is the President of Numertus Research. Before the break, Johna, you were in a way challenging the way outsourcing has worked for most people. Unless otherwise, it is very tightly handled. He could go on a tangent and there’s a lot of money locked and that question was to be completed and then Wallace, you were to respond.

Sanjog Aul [00:27:55]:
So Johna, you want to just finish up or we repeat?

Johna Johnson [00:27:58]:
Sure, my specific question, Wallace was really more along the line, but how do you keep the business interests of General Motors aligned with the business interests of your suppliers, your outsourcers? In other words, are you sharing metrics and then somehow rewarding them if they help you metrics or okay, I couldn’t if we’re punishing them if you don’t. How does that work?

Wallace Dalrymple [00:28:18]:
With all of our suppliers and our partners, we have SLAs in place that address how things are negotiated from a metric standpoint and how those are that whole risk versus reward and if they meet their metrics or not and we have the foster inflation scorecards and very close relationships with our vendors but one point that I think is really one of the the key things for General Motors in its outsourced model is standardized work because of the standardized work processes that we have implemented from an ISMS perspective, those standardized work processes are also shared with all of our suppliers and all of our vendors. So let’s say for instance, that they, they can’t work with those standardized processes or there’s an issues or concerns with that particular supplier or vendor. We have a capability because we’ve implemented standardized work to move suppliers in and out as needed or as the business needs fit. So that underlying foundation is really what has helped enable us to be able to work with many different suppliers or change out suppliers.

Johna Johnson [00:29:24]:
So if I understood you correctly, and you do correct me if I’m wrong, what you were, what you’re really highlighting are two key success factors for you have been the ability to define business metrics and hold your partners to it and the second one is to really standardize the work process so that there’s very little confusion or ambiguity as requests are handed off. Is that exact? Did I get it right?

Wallace Dalrymple [00:29:45]:
Yes, you did that. It’s 100% correct.

Johna Johnson [00:29:48]:
Okay and the reason I talked about the business metrics by the way and there was a very interesting conversation I had with a large airline company that had successfully outsourced a lot of its IT. I believe all of its IT and this is not a US Airline by the way and their, and their response? Well, actually what they did from a business metric standpoint is they simply rewarded their outsourcer for number of planes that arrived on time and punished the outsourcer if the plane didn’t make it on time. It was that simple. It wasn’t. They had no technical level metrics in there.

Johna Johnson [00:30:23]:
I’m sure they defined the technologies they were looking for and managed to that but at the end of the day the business alignment was such that both the outsourcer and the airline cared about the same fundamental operational metrics which I found really interesting. Although I would say that not every airline in the state ended that model but I just thought it was a very interesting way to handle your outsourcer.

Wallace Dalrymple [00:30:45]:
I agree with you. The business and the business organization and the IT organization, the metrics that are measured really are dependent on each other. So if your IT metrics as far as your FTLA’s are concerned, which we could date uptime and depending on whether that’s three nines, four nines or five nines if in like the General Motors environment, manufacturing, if we don’t have that availability to build cars and trucks, then that’s an impact to our business. So they have, the business is involved and it has to be in order for us to meet our customers objectives. Now let’s talk about the fact that.

Sanjog Aul [00:31:22]:
There’s a lot of mergers and acquisitions going on and we are trying to already having a lot of issues with handling infrastructure and trying to make the agile and just keep it running for that matter and now this, a big disruption comes. So how is that planned for from an infrastructure management standpoint so that you can become agile or if you are agile and you had planned for everything and suddenly there’s a disruption. So how do you take it back to where it was?

Johna Johnson [00:31:49]:
I think Wallace wants to answer that, right Wallace.

Wallace Dalrymple [00:31:51]:
Within General Motors we have many subsidiaries in many business areas. GMAC and Ontar and then Delphi used to be part of General Motors and that was spun off. So we’ve have practice with how to either bring on new businesses into the General Motors organization or to separate them and I go back to earlier in our conversation when we were talking about decoupling. Your application and your infrastructure have the ability to be decoupled so that there is a dependency, but your business processes aren’t dependent on that infrastructure tied to your application, then that makes the separation a little bit easier. So now you’re just focusing more on network infrastructure as far as maybe your T1 and maybe some of your switches and your routers and maybe configuration changes within those but from an application perspective, the impact shouldn’t be there.

Johna Johnson [00:32:50]:
And I would agree with jokingly hunting there, but I think that, decoupling is a good piece of this. I think the second big piece of it again is to have both second and third tight metrics and an exit strategy. Don’t assume that since you’ve thrown all your eggs in the basket of your provider, whether an outsourcer or just a general strategic vendor, that you are somehow absolved of the responsibility to also have contingency plans for that company to go if they go belly up or get bought or what have you and the one thing that I would also correct on that, quite frankly, is that as part of the procurement process, one of the things we recommend all the time with the folks that we work with is that they do a business analysis of the vendors that they’re working with and really try to understand what the chances are of that vendor getting acquired or acquiring someone else and getting distracted or going out of business and historically, at least back before the boom, there was a sense that you only really need to do this on very small companies if you were taking a great risk because you were going with a startup and now I think people have recognized that even very large companies such as, well, L&T can get acquired and so you need to really think through that and do your business analysis on very large companies as well as small companies and not nut for doing that. It’s an important part of the procurement of that firm.

Wallace Dalrymple [00:34:08]:
I think you’re correct on that and seeing with the telecommunication, with the consolidation of all the telecommunication carriers, I know that our business expectation wasn’t that there would be a disruption in service. When an AT&T or an SBC get together or whoever that telecom provider might be, the customer expectation, there won’t be an interruption in service and hopefully that doesn’t happen and in our case it hasn’t so far so.

Sanjog Aul [00:34:34]:
So one thing we talk about is M&A among vendors and outside entities. When we have M&A, is that any different and we can always answer this question after Johna, you finished your thought?

Johna Johnson [00:34:44]:
Oh no, I would like for the thought.

Sanjog Aul [00:34:48]:
Okay, so what basically what I meant is like when we spoke about like the telecom M&A, that happens and that is totally understandable. That is a business which is user of that particular service should not even be thinking about, hey, I’m going to have suddenly an issue with an on demand scaling that you are expecting from that vendor but if there’s an M&A going on within our own company, what are the typical things that a company would do as a part of planning process and that is specifically for infrastructure because other things could potentially wait but infrastructure has got a zero tolerance.

Johna Johnson [00:35:19]:
I think there are a couple of things. You mean the part where say a cham goes out and buys socks, for example, that kind of M&A. I think one of the things to think about is to actually have an accurate and as at all possible automated and updated inventory process. You would be amazed how many companies butcher the task of asset management. If you don’t know what you have and it’s true, whether it’s anything as simple as desktop or even circuits or servers, a lot of the infrastructure you probably don’t know what you got. So step one is to make sure you do know what you have. Step two is, since you’re moving towards an agile infrastructure anyway, one of the first steps from that path is to standardize,

Johna Johnson [00:35:58]:
so you should, that doesn’t mean you always use the exact same profile, exact same configuration for everything but you should at least understand when you’re going to deviate from the standard architecture for something or what the standard architectures may be. So that way you can minimize the number of different configurations that you’re supporting at any one time and then the advantage there is when you do a merger or an acquisition, if you’re spinning something off or acquiring a whole new company, first of all you understand the process for getting from your hodgepodge mismatch of infrastructure to a standardized one and you can immediately start the new acquisition down that path. The second thing that you can then do is handle if they’ve already gone down that path and standardized, even if they’re the configuration they standardized is different from the one you standardized on. Do at least have a process in place for managing different standards and you can immediately tackle that. I’d say there are a lot of little nitty gritty gotchas.

Johna Johnson [00:36:54]:
For example the degree to which you can get communications integrated. Things like for example your address basis can be challenging. If everybody’s using a net 10 address space and you buy a company that’s another net 10 address space and all of a sudden you have to have firewalls between you and them just to do the network address translation. There’s lots of little nitty gotchas, but if you’ve got an accurate up to date net inventory of your assets and you have a well thought out standardization plan, you can really smooth a lot

Johna Johnson [00:37:23]:
of it, a lot of the hurdles.

Sanjog Aul [00:37:25]:
Before we get into your response, Wallace, let’s take a very quick break and we’ll be right back after these messages.

Sanjog Aul [00:39:57]:
Welcome back to the show folks, For listeners who just tuned in, today’s topic is Agile Infrastructure Is It An Oxymoron? Our guest for today’s show are Wallace Dalrymple and Johna Johnson. Wallace is the Chief Network and Telecom Architect, Global Technology Management with General Motors and Johna is the President of Nemertis Research. Before the break we were talking about how an organization can get impacted when there’s an M&A and what kind of disruptions occur and at the same time, for example, for the same company. The example that Johna used was VM acquiring and what, what happens when you have got an agile infrastructure here or expectation of having an agile infrastructure in one company and the other one and then one merges with the other. So how’s that transition taken care of and how do you come back to the same state if at all you were already agile while if you were about to share your couch.

Wallace Dalrymple [00:40:46]:
Yes, thank you. In mergers and acquisitions it would be an understatement to say that’s no easy task but with what we’ve done at General Motors is with strategic planning, standardized work processes and using the idle process framework that we have today and looking at how we would actually bring those organizations into our environment and identify. You mentioned, Johna, you mentioned earlier about asset management, which is obviously key to knowing what you have or what you’re taking over. If I could add on to that, we would also look at things as far as the availability of the systems and taking over more of a service level management, new help desk processes, problem management processes and handling any new incidences. So the obvious goal is that there would be no impact to either environment and trying to bring those technologies together, even an overnight solution and there really has to be a really thought out plan on keeping the existing applications up and running for either an organization that has merged with you or that you’ve acquired and keeping business continuity up and running. So asset management, business continuity, capacity management, availability management.

Wallace Dalrymple [00:42:02]:
I think those are just some of the components that would go into that whole strategic planning initiative.

Sanjog Aul [00:42:09]:
Talking about standards and frameworks, we definitely know the new wave of Idle which has been around for a while and everybody’s joining that bandwagon is that the only thing that people should be looking at while trying to standardize what they do, how they do or and also at the same time what kind of amendments that they would want to make to this particular set of standards or frameworks where if they want to go and become agile in terms of infrastructure.

Wallace Dalrymple [00:42:36]:
Johna, I’d like to start with this one. Idle is not an end all solution and it’s just a framework. It’s a process framework that companies can leverage to build on those Novon processes, IT service processes as a starting point. No, those processes should be tailored to whatever that organization’s business model is but it just, it’s like any other standardized process. It gives you a foundation to start from and then you make those changes as you see that some of you might use some of them, you might not use journal. What are your thoughts on that?

Johna Johnson [00:43:12]:
I couldn’t actually add to that because I think you said it perfectly. Don’t assume Idle is going to solve everything. Use it as a baseline starting point and then build from it and standardize from it as you see fit. I do have a question for you though, because this is an interesting one. I’ve seen with a lot of the folks I’ve worked with that have gone through significant mergers and acquisitions and it’s a real challenge for them in a lot of cases. How do you make the decision when, let’s say they’re Oracle and your SAP. How do you make the call as to which system gets test or whether or not you’re going to continue supporting those systems? I don’t know if you run into that situation.

Johna Johnson [00:43:50]:
Another great example is there on Sprint you’re on MPI or whatever. How do you make that call or do you make that call?

Wallace Dalrymple [00:43:58]:
I think you have to definitely address it and to make that call there’s many different factors. You have to look into what makes the most business sense for your organization. There might be costs associated, there might be some synergies in some economies of scale by maybe merging the applications if possible. You mentioned Oracle and SAP, which are two separate applications that don’t have that ability to just use one or the other because there’s a lot

Wallace Dalrymple [00:44:26]:
of customization that goes on with that

Wallace Dalrymple [00:44:29]:
but from day one, I don’t think that you could just say you go to one or the other because you got to keep those business processes up and running in the interfaces to those applications as well is very important. So I think the longer term strategic plan that you have to come up with because there’s no way just to switch from even going let’s say better example might be SAP and PeopleSoft two distinct applications with distinct business processes that migrating from one to the other is more of a strategic plan.

Johna Johnson [00:45:00]:
That was actually what I meant when I said Oracle. They’re now Oracle. It’s all Oracle now. They have to see Larry Ellison for everything.

Sanjog Aul [00:45:11]:
Now one last question I have for you is when we talk about all the stuff that we need to do at the shop floor, if you will, in terms of getting the infrastructure straight and everything, what is the typical grooming required for both the leadership so that when you go sell to them for more money for putting it into this making your infrastructure agile and better and also other hand the infrastructure professionals who are actually doing the job so that they really understand the big picture and put that individual brick to make that big palace that you always are dreaming of.

Johna Johnson [00:45:41]:
So I guess if I understand the question and maybe let me point out exactly, but what are you getting at here?

Sanjog Aul [00:45:49]:
So basically when we are trying to groom over a leadership or prepare not exactly groom or prepare our leadership and groom our infrastructure professional to essentially while we are making all the investments in to make the infrastructure agile, there is some the people side which we are talking about what kind of grooming are we talking for infrastructure professional, what are we preparing them and as well as the leadership because the transition is what make people scare.

Johna Johnson [00:46:13]:
I think, see the reason I was a little confused is I think there’s two separate problems bundled into this. One is the standard issue problem which is how do you sell leaders on infrastructure and the answer is you don’t because it’s like trying to sell an apartment building or facility based on the quality of for example, the bricks in it. You don’t, you just say it’s a really beautiful apartment and look at the wonderful engineering that’s gone into it and everybody knows that living in a well engineering building mean that you aren’t going to get sudden leaks in the plumbing and the power is going to stay there. If you count the benefits that it’s,

Johna Johnson [00:46:49]:
the benefits to you are that it’s pretty, it’s attractive, it’s convenient to live in. There’s not a lot of unscheduled outages or downtime and that’s really what you pitch. You pitch the benefits of the infrastructure, not the infrastructure itself. You don’t go into a loving detail about the quality of the brick and the microscopic configuration surface and the met, even though that may be a critical component of why the building is so permanent to live in, but that’s thing one. So just coming back to some of the examples we’ve used here. You don’t talk when you’re pitching a virtualized infrastructure to enable SLA. You talk about the rapid development timeframe that SLA gives you and the fact that it needs this infrastructure in order to do that.

Johna Johnson [00:47:30]:
So that’s the general answer for how do you sell higher up on the value of this agile infrastructure so that they invest in it. I think the more interesting question though, is the one that I wanted to sort of understand completely, which is how do you train people? How do you deal? What are the skill sets that are required to class A, an agile infrastructure? I think that the answer is a little bit of a puzzle because in some respects you need more people who are generalists and fewer people who are specialists. It’s just what are you specialists and let me talk about what I mean by that. If, for example, you’re making a strategic initiative to outsource and effectively you’re going to want people who understand a broad range of technologies rather than whole battalion of people who are incredibly deep in a particular technology because understanding how technologies fit together is key to managing the outsources, who are providing services that multiple technologies, for example. However, in certain cases, you really want to have people respectful from their technology and an example would be storage. One of the successors we found with enterprises that succeed in actively virtualizing their storage and creating a very agile storage infrastructure is that they actually have a storage tag team that is, even if it’s a small group, and it usually is, it’s under 10

Johna Johnson [00:48:51]:
but people who handle storage enterprise wide, rather than having storage being built in in the shadow of the servers and desktop and mainframe architectures. Once you start thinking as an organization about storage enterprise wide, that’s when you can actually make real progress and, doing standardization and virtualization. So the answer is simultaneously you’re looking for people who have a broad expertise in technology, but you need to also focus on the key, some of the key technologies that you’re relying heavily on in order to enable this virtualized and infrastructure roll.

Wallace Dalrymple [00:49:27]:
I agree with you and a couple points in total, I think everybody understands that the infrastructure and an agile infrastructure is a business enabler. So if IT is to be effective, it needs to meet the maintenance objectives and those business needs and one quick example that I can give of that is we all go to the rental companies at the rent vehicles when we get off the airport or go to a city what a lot of what the rental companies, most of them have done is to leverage IT. It used to be you have to go you had to go to the desk when you bought your car back and you had to sign up day fork and get it back. Well now they actually give you that information and you check you out right when you pull up the car. So they have leveraged infrastructure and made it more agile to meet not only the business need for the rental company but also improve customer satisfaction and to get back to Johna, the business skills. You’re right, that is a puzzle and I think business skills are harder to come by than IT skills from the ability of being technical versus understanding what your business needs.

Sanjog Aul [00:50:39]:
On behalf of the show and our listeners, I would really like to thank you Wallace and Johna for sharing your thoughts and insights about how organizations can meet IT infrastructure management related challenges and actually create an agile infrastructure.

Wallace Dalrymple [00:50:54]:
Thank you.

Johna Johnson [00:50:55]:
Thank you for inviting us. We appreciate the opportunity.

Sanjog Aul [00:50:58]:
Thank you again.

Sanjog Aul [00:50:59]:
Now IT infrastructure management surely comes with a set of challenges as we all saw, but at the same time what we can do is to use a good mix of people, processes and technology as again discussed during the show, and try to make it actually happen. What are your challenges while trying to make your IT infrastructure agile? Please send us your story at views.ciotalkradio.com that is views.ciotalkradio.com thank you again for listening to CIO Talk Radio. This is Sanjog Aul, your talk show host. Till next week take care and god bless.

Download Podcast
Apple Podcast, Google Podcast, Spotify, Pandora, iHeartRadio, SoundCloud, TuneIn, and Stitcher. Find other syndication channels here or search CIO Talk Network podcast on any other app.
Explore More

• Challenges of a Large IT Implementation
• Redefining Enterprise Architecture
• IT Infrastructure Library (ITIL) – Is it a Silver Bullet?
• Successful SOA Adoption: What does it Take?
• Agile Infrastructure: Is it an Oxymoron?
• Is your IT SOA (Service Oriented Architecture) Ready?
• ITSM and SOA: Siamese Twins
• Rewiring Your Enterprise Architecture With Social Media
• What Would it Take For “Collective Intelligence” to Become a Reality?