Prescription for 24/7 Uptime | Article

metallic clockYears ago, many organizations had software systems that were “nice to have.” In today’s world, though, we need systems to help us compete in the marketplace more effectively. Software systems now have a major impact on our business and enable us to produce higher revenues, higher income, and higher market capitalization. When a system goes down, an organization is no longer competing on level ground and can lose customers and millions of dollars in revenue.

Not too long ago it was very common for systems to go down and an SE to be dispatched to bring the system back up. People were accustomed to the fact that that was the way things worked. “We have raised the bar and now have higher requirements,” comments Thomas Jones, CEO of StrataSource. “People these days want their computer systems to be as available as their telephones.”

The Plague of Downtime

Jones says that neither the operating systems nor the data bases themselves cause the problem that results in downtime. It is the end user’s interaction with them that can cause a system failure. “We see companies doing their own database administration without the necessary expertise,” he says. “They do things with the data base that doesn’t bring the system down today but, in a few days, will cause bigger problems that eventually will bring the system down.” He explains the problem further by saying that all the automation in a company (desktop, marketing, sales, support, manfacturing, engineering systems) goes through the network to a server. At that point, the automation breaks down and becomes a manual process because there is an individual who oversees the administration of it. That individual is the weak point, and the only way to strengthen the process is to automate it from end to end.

Downtime is costly, and end users seek ways to ensure that their applications will be available. Additionally, most companies now focus on their core competencies and try to outsource the functions that are not core. Although data base administration is not a core competency for most companies, it is a function they perform because there has been no alternative–“unless you are a DuPont or a Kodak who can bring in an EDS or an Arthur Andersen to do it for you,” he explains.

A Doctor Who Specializes in Maintaining Uptime

StrataSource is a spin off from a consulting company called Relational Data Systems. That company performed customized application work at customers’ premises. Before long, the company founders realized their need for remote data base administration to help those customers. They developed the concept that is now the product called Lights Out. When they realized it was the star of their offerings, they spun it off and formed a new company.

That company is StrataSource, a remote data base administrator. The software monitors an end user’s infrastructure, the underlying data base and operating system on a 24-hour basis. When an issue arises, the software sends a message to StrataSource’s operating center in Irvine, California, detailing what the problem is. StrataSource personnel then log in to the system and fix the problem.
In today’s environment, Jones explains, it is not acceptable to manage infrastructure in a reactive manner. By managing proactively, problems can be fixed while still small and will not be catastrophic in the future. Jones compares the management methodologies to an analogy from the medical field. “Current data base administration is done is much like what you run into in an emergency room,” he explains. “If you are in a car wreck and go to the emergency room, the doctor will look at you and realize your system is down. To bring your system back up, he has to stop the bleeding here, fix a broken bone there. In this reactionary mode, the doctor still doesn’t know what all is wrong.”

With StrataSource, the software placed on end user computer systems proactively monitors the underlying data base. A reported problem can be fixed when it is still minor. “Let’s say a doctor puts a monitor on you and then studies the reports and informs you that your enzyme levels are a little high and that this might cause you to come down with appendicitis,” he explains. “So he treats the problem with drugs. If that works, fine; if not, something more major might be required. In this way, you stay healthy in the process (your system stays up).”

StrataSource actually fixes problems before its customers even know they exist. An example of this service capability is when one of the support managers in the operating center saw a report alerting them to a customer’s problem. It was 7:00 PM at the center in California. He fixed the problem and sent a message to the customer informing them of the corrected problem. At the customer’s location on the east coast, it was 10:00 PM. No one was there. StrataSource automatically gives 24-hour service; customers select whether they want that coverage for five or seven days per week. “A few customers have asked to have the service for just eight hours,” he says. “They really did not understand the issue. You can have as many problems when you are not there as you do when you are there.”

Besides increased uptime, the benefits of using StrataSource to maintain a healthy system include solving the people problem. “Most end users have a difficult time finding the data base administration people they need with the required skill levels,” Jones explains. “Years ago, you just had one data base and one administrator to work on that. Now companies bring in many new applications, each with its own data base. One person does not have all those skills. There are not enough people now; yet we are installing more systems. There are few people coming out of the universities now with a career ambition to be a data base administator.” Along with downtime, the people problem is a costly one.”If you want to have somebody on three shifts, it could cost you anywhere from 1/4-1/2 million dollars just for data base administration,” states Jones.

From Concept to Customer

StrataSource markets its service through independent application software vendors. Jones explains that the vendors also reap benefits from this service. “Often they sell applications into environments that are not necessarily ready to support the applications,” he says. “That causes them a lot of cost and headache in the future because they will get calls from customers whose infrastructure isn’t properly managed. Our service reduces the vendors’ ongoing support costs because they will only get calls about their application instead of the infrastructure.” Offering this service with the applications also gives the vendors a way to differentiate themselves in the marketplace.

Once a client signs up for StrataSource’s service, the company audits and certifies the client’s system. The application vendor then loads the StrataSource software with the application. The software is passive and occupies little space on a system. It takes only 48-72 hours to get a client’s system up and running with this service. Customers sign annual or multi-year service contracts and pay up front.

Incorporated in January 1999, StrataSource already has 45 customers (and supports 65 sites). Some customers were brought over from the former company, and some are the result of word of mouth. The company provides service now in five countries and is looking at taking its service to Japan and Europe, where service is a key requirement.

“This is a great new field,” comments Jones. “We are pioneers but, as we talk to industry leaders, end users, and analysts, it is obvious we have hit on a real hot button that has been overlooked. Everybody has been looking at the other end of the equation–how to get the end user new screens and new applications. They forget that they never upgraded the process of managing the infrastructure of these computers.”


Post a Comment

Your email address will not be published.

( required )

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>