July Issue of the Taos Newsletter: Systems Management and Remote Management

Interview with Jeff Lucchesi, Senior Vice President of Worldwide Operations, Corio, Inc.

Taos: Tell us a little bit about your company.

Jeff: Corio has been around since the late 90’s and started as a consulting company. In 1999, Corio changed it’s name and became an applications service provider (ASP), providing an integrated suite of professional services, applications management and infrastructure services within a flexible, pay-as-you-go, utility-like cost model we call Applications on Demand TM. We’re actually the first company to use the “on demand” terminology.

We always saw software as a service and understood the need for “pay as you go” type of services. Really, that’s the whole background of ASP’s - pay a certain fee, pay for what you use - like Salesforce.com, which many people know of as a result of their recent IPO. ASPs provide a hosting service and you pay so much per seat and it goes up and down depending on how you use it. Customers like that model because of flexibility. We don’t really go in and take a complete outsourcing stance with our customers. It’s more of a specific focus and our focus is our ERP (Enterprise Resource Planning) and CRM (Customer Relationship Management) applications. Our goal is to work in tandem and become partners with our client CIOs and their whole organizations.

It works very well. We’re used to good working relationships. Last year, we hit 68 million dollars. We’ve grown every year, even through the tough times.

Taos: Tell us about how you use remote management, both in terms of managing your own infrastructure as well as managing the service that you provide to your customers.

Jeff: One of the biggest reasons customers use our services is that we have some really good management tools and we do a lot of management of servers and networks for our customers now. In this area, we have a couple of offerings:

• One is Corio Full Service TM, which means we run the systems in house. We do all the client backups and systems administration as well as manage the application. We do all of it - it’s a “soup to nuts” service.
  
• The second is called Corio Onsite TM which is targeted at larger customers who already have investments in datacenters with a lot of equipment and a large operations group. They don’t need for us to manage their assets. What they don’t have is maybe expertise around PeopleSoft or SAP, so they’ll have us manage their applications remotely.

Regardless of which option the customer chooses, what’s very attractive to them are the monitoring and management tools we use. We really go deep into the application management side - we get into jobs that are processing, we get into certain other processes of the systems that may be having problems or issues, etc. A lot of our value is in being proactive so that PC issues on the database are usually seen by us before the customer and handled before they experience the problem.

A simple way of looking at it is like the warning lights and gauges on the dashboard of your car. For example, you can see that you are low on gas - you’re not out of gas, but you know you are close. The same holds true with management tools. They indicate certain conditions are existing on the system, whether on the application, or in the database, the server, etc. that you should probably take care of before they get to a certain point.

We have alerts that are built into the system that say “if you get this, tell me” and we have it all automated so that it gets routed to the correct person. Here are some examples:

• A database alert that goes right to the database group.
  
• A potential spike on the CPU goes to our SysAdmins.
  
• Maybe the customer is not experiencing any problems, but if the behavior is something odd, we’re notified.

Taos: And is this a homegrown system that you use?

Jeff: It’s a combination of scripting we’ve done and some other tools like BMC Software (www.bmc.com). We use about three or four other tool sets that tell us certain things are going on - some transaction monitoring tools - but mostly BMC at the core of it. There’s a tool we use that takes a lot of data and filters certain pieces of information out. It’s a knowledge manager - it says “if these conditions exist, then you have a problem, or if these conditions exist, you may be having a problem so take this ticket and route it.” We just found ways to enhance it. We also use HP Openview for the networking.

In total, we’ve made a very large investment in tools and in the people to support those tools. We also have a major investment in security tools. In fact, it’s one of my biggest investments. As a result of our tools capability, some customers will buy almost a menu of products from us. They’ll say “Give me the monitoring capability. I want you guys to do that. And I also want your security package because that’s a lot more robust than what we have and what we envisioned.”

All of this is really new capability and very useful. Just a few years ago we had a big monitor in the center of the room with people looking at screens around the clock. Lights would go off and someone would have to drill down to find out what was happening underneath. We’ve automated all of that.

Taos: How much customization did you have to do to the tools that you use?

Jeff: A good bit, but it’s well supported by the vendors. There are knowledge modules that you use to make customizations and set up rules. There is some scripting that we have done to enhance some of that capability, but each tool generally comes with a lot on its own.

Taos: How did you come up with the strategy that you built for remote management and the tools that you use for it?

Jeff: It’s market-driven – it’s what the demand is and what the customers want. And they want more all the time. So, we’re constantly looking at the available tools out there and asking ourselves “What’s the next level that our customers are going to demand?” We find it’s usually at the application layer. The infrastructure is pretty well managed by most tools now and is not rocket science anymore. Now the questions customers are asking are “What’s happening inside my application? What’s happening with the transaction?” Or it’s more business ruled. For example, “I can’t afford to have a manufacturing error during month end, so what are you going to do to make sure that never comes down?” This means they want certain transactions monitored. Customers are getting very specific about what they want monitored and how. Monitoring is hot right now.

Taos: In terms of what is available to provide - the be all and end all of monitoring and remote management - what’s lacking? What aren’t you able to do today that you’re expecting is on the horizon, that you should be able to do?

Jeff: Right now we’re meeting the demands that we’re getting. However, we’re pursuing tools to allow us to get into more intricate details about what is happening in the application. For example, at the manufacturing transaction level I just mentioned, there are some tools that look at the application more from that perspective, but they’re pretty early in the development process and not quite where they need to be. There may be one out there that I haven’t seen yet, but from what I can tell, what’s out there is still fairly early in their financing or their development.

The bottom line these days is that the majority of issues are either in the database or the application. And many of them relate to customizations customers have done to the application - maybe they made a change last night; all of a sudden it’s not behaving. You need some way to quickly deduce this and get to the heart of the matter.

Taos: Do you think that in the future there will be a common industry standard for remote management?

Jeff: I think some of the underpinnings will be with the products. I think the whole area of remote management has a long way to go. There’s a lot of opportunity and I don’t yet see a single leader emerging out there - someone who owns the market on remote management. In fact, right now, with all the things we are monitoring, I’m not sure I can see any one company take over and say they’re the leader of remote management.

Taos: And to the extent that there is anyone out there that’s capturing this space, you’d say it’s BMC?

Jeff: Well, BMC really has the infrastructure - the performance level, and has done a pretty good job getting into the application layer. We use BMC for PeopleSoft and Oracle, but we also use a product called RealTec for SAP. It’s really an SAP-specific tool which makes it very good at monitoring SAP. And that’s what our SAP customers have been demanding, so we have shifted our focus for SAP to RealTec.

Taos: What do you want to see as the next step in this space?

Jeff: In the next generation, I want to see a product that isn’t an agent on the system. Instead, I’d like to see one that is more of a remote product that will give the data required. Currently, there are still a lot of agent type products that have to be loaded onto the servers. And since I have over eleven hundred servers, that’s a pain.

Taos: Do you know that to be something the vendors are working on?

Jeff: Yes.

Taos: In the future, do you believe the majority of hardware will actually have support for remote management built in?

Jeff: I think to some degree yes. Currently, some of the applications already have incorporated it in some form. SAP has some management of its applications built in. I think you’ll see it at the application layer. You may also see it at the operating system layer.

Taos: What functions and assets are still managed better locally? Why does anyone have to be physically in touch with their systems?

Jeff: I still hear customers say “I want my server and this data center because I can go down and see my server.” It’s mostly for comfort, but other than that, I don’t really see the need for any of my customers to be close to their systems. Networking is pretty powerful. A lot of the systems we manage are several thousand miles away - either in Bangalore, India or in San Carlos, Phoenix, Chicago, or Dallas. Systems can be supported anywhere in the world.

Taos: Obviously, there is a lot of benefit in staff performance and productivity if they don’t have to get up from their desk to manage the system. Are there concerns on the opposite side with regard to latency, especially the further away they get from the system?

Jeff: Some of the tools do have a latency issue over a VPN. But, in most circumstances, there are workarounds for those (except for some third world infrastructure challenges). So they really haven’t inhibited us from providing service.

Taos: Given all the benefits to remote management, what’s your take on why people still feel their data center needs to be local or at least well staffed beyond the remote hands?

Jeff: I think it’s purely an emotional argument - I don’t see any logic as to why the server needs to be nearby. Even if the data center is only ten miles away and you’re in a Co-lo, you still can’t just walk in and play with your server – access controls and things of that nature. In those cases, there are still remote hands that can reboot your server, and they don’t need to be sophisticated. So you don’t need to get in your car, drive ten miles, and throw the switch. It’s kind of like power plants. I don’t need to know that my PG&E power plant is two miles from me. I don’t care, as long as I have power, right? Same with servers.

Taos: Do you foresee a time in the very near future where data centers will be in, say, China and you’re doing remote management from here and, say, India?

Jeff: Actually we are working on a deal right now with a company whose datacenter is in Singapore. Singapore’s networking is pretty solid, but they want our India group to manage it. My US operation would back up India. Those requirements are definitely out there. For those countries that don’t yet have the infrastructure Singapore and India do, it’s just a matter of time.

Taos: Do you foresee any challenges for that in terms of the link?

Jeff: That could be an issue but, for us, there's always somebody in the US and always somebody in our Bangalore office. Should one of those links break down, we have enough redundancy to cover it with no problems. It’s part of our disaster recovery plan. We always have staff in two locations that can manage problems. We don’t put all our eggs in one basket.

Taos: How often should remote facilities be audited?

Jeff: We do two a year, because we do a physical inventory of our equipment twice a year. We always go through our facilities and make sure everything is sound - the cabling is sound; the physical layer is working.

Taos: Obviously, a lot your strategy around remote management has been necessitated by your business. It’s core to your business. There are a number of companies out there who are co-located and who have their data center at a distance from them, even if it’s in the same building, looking to do more remote management. What advice would you give them for how to evaluate what level of remote management they should put in place and how to build their strategy?

Jeff: A lot depends on what they’re doing. They have to do their own risk assessment and figure out what business they’re in. They need to ask questions such as:

• What are our priorities?
• What needs to be up and running?
• What can I live without if it’s down for eight hours or twelve hours or two days?
• Do I really need HR up every hour?
• Do I need manufacturing up all the time?
• What do we need to do to support our Sarbanes-Oxley requirements?

There are a whole series of things they should have on a checklist that helps them build their own internal priorities and SLAs (Service Level Agreements). Very importantly, the business units have to evaluate it. They provide a service to the business and they need to be able to respond to what the business says is needed.

From there, they can determine what they need to monitor, how they need to manage that monitoring, how they need to respond to issues, etc. Then, if they want to do it remotely, there are tools available like the ones we use that will allow them to do that.

For companies that can’t afford expensive vendor systems, there are some good open source solutions out there that we’re also looking at right now.

Jeff’s mission as Sr. Vice President is to drive customer satisfaction for Corio’s customers and streamline Corio’s operations processes to optimize profitability.

Jeff brings to Corio a successful career of planning, implementing and managing IT solutions. Jeff has over 20 years of experience in key IT management positions for companies in Silicon Valley. Immediately prior to Corio, Jeff was Chief Information Officer at DHL Airways where he managed a team of 120 developers and architects and was accountable for operations, applications development, field engineering, and business process reengineering in the U.S. During his tenure at DHL, Jeff held a number of management roles in IT and achieved key accomplishments including significantly reducing telecommunications costs and implementing and supporting new web-based applications and services for customers. In addition to these executive positions, Jeff held MIS, engineering and management roles at Syntex Corporation and Syva Company.