Business Analytics (Tivoli)

The Business Challenge:

When a patient is rushed into a hospital emergency room with an unknown illness, doctors don’t have much time to think. They rely on advanced diagnostic tools to get to the source of the patient’s ailment quickly, so they can apply appropriate remedies.

Likewise, when there’s a glitch in a hospital’s clinical or business-critical applications, all eyes are on IT. And, like the doctors in ER, the IT staff does not have the luxury of launching a time-consuming investigation.

Darryl Shorts, a senior IT engineer and architect with the Franciscan Missionaries of Our Lady Health System (FMOLHS), knows that all too well. A 12-year veteran of the Baton Rouge, Louisiana-based FMOLHS, he’s seen frustrated IT staff having to log into multiple systems, generate reports, then interpret and correlate them to find out what’s impacting performance. “We were chasing needles in a haystack until we could get better visibility and correlation,” he says.

Needless to say, FMOLHS’s 1,400 medical staffers, and more than 7,500 administrators and other personnel were feeling the pain when systems slowed and the cause wasn’t readily apparent. That prompted upper level IT management to seek better insight into systems performance.

The Solution:

To reduce diagnostic time and meet the reporting needs of the executives, Shorts asked Mainline to help implement more effective business analytics. Because FMOLHS is an existing Tivoli Storage Manager (TSM) and IBM Tivoli Storage Productivity Center (TPC) customer, Mainline was able to leverage the Tivoli Common Reporting (TCR) platform embedded in these products to achieve the goal.

The opportunity is made possible by the fact that TCR is based on IBM Cognos, which is a best-of-breed business analytics framework known globally for reporting, dashboarding and scorecarding capabilities. The engagement required no additional licensing, simply the addition of a few weeks of consulting services to extend Tivoli Common Reporting to make it easier to create consistent, meaningful reports and share those reports across the enterprise.

Even though FMOLHS was a longtime Mainline customer, Shorts still did his due diligence. “We sent requests for information to multiple companies,” he recalls. “Mainline was the only one that had the expertise to even start a conversation about Cognos capabilities with TPC and TSM at the time.”

Mainline helped FMOLHS rework its Tivoli reporting workflow to create aggregated application-level dashboards. These contain custom metrics and thresholds that drill through by passing parameters to existing detailed granular reports.

Typical reports are often oriented towards infrastructure management, but this new model starts at the business level and, with a few clicks, can reveal underlying technical issues. The first level of reporting for each application is a scorecard, listing metrics that FMOLHS considers pertinent to each application. Colored lights show performance status: green means the application is meeting its service-level agreement (SLA). Yellow means performance is marginal; red shows failure to meet the SLA. For non-technical executives, this may be enough to stay abreast of the situation.

Admins can drill down into TPC and TSM from each of these indicators to pinpoint the source of any performance problems. With Tivoli Common Reporting, both TPC and TSM are accessible through a common portal.

According to Shorts, after Mainline’s services engagement Tivoli Common Reporting is a much more user-friendly way to understand what’s going on with an application. “Our application admins understand their applications, but not necessarily the infrastructure behind them,” he says. “With the Tivoli Common Reporting dashboards and scorecards, they can see if they have anything yellow, and what the correlation is from a system perspective.”

It took less than a month for Mainline to develop scorecards and dashboards for applications such as payroll, materials, general ledger, accounting and other modules in FMOLHS’s Infor ERP system. “Later, we’re going to add system performance metrics such as CPU, time, memory used and available, page file utilization and ultimately, transaction time,” Shorts says.

The Result:

Faster remediation
With customized dashboards and easy-to-read scorecards, FMOLHS IT admins and managers have greater visibility into their systems and gain insight into problems sooner, leading to faster remediation. Without this capability, it would take two admins each in three IT groups (storage systems, storage network, and operating systems) 20 to 30 minutes to log into the different management tools and get the initial data. Then the groups would have to compare data to discover the source of the issue. Now, they just look at a screen and drill down. “We’ve literally gone from about an hour and a half for this process to a couple of minutes,” Shorts says.

Better resource utilization
With IT anomalies arising typically three times a month, FMOLHS is on track to save more than 200 IT person/hours every year.* This is time that the IT staff can reclaim for more valuable projects-an important consideration for a 25-person team that’s responsible for information systems on thousands of servers in four large data centers.

Higher value to the organization
With the time saved and the relatively small investment in professional services, Shorts estimates 100 percent payback on the initial engagement in about three to four months. But he’s far from finished. “It came out much better than we expected,” he says. “That’s why we got funding for additional work.” FMOLHS is now engaging with Mainline to add functionality to the Tivoli Common Reporting dashboards and scorecards. “We want to include clinical systems such as Spacelabs Healthcare and our registration system,” Shorts adds. “They are highly visible and critical to the operation of the enterprise.”

The success of this analytics project may have garnered a lot of attention from the C-suite, but on other fronts, Shorts is happier to be ignored. “We’ve got teams that can completely bypass us at this point,” he says. ”They don’t have to call us and say, ‘What’s going on with the system?’ They can just go over and look.”

* More than 1 hour saved per incident × 6 people × estimated 3 incidents per month = 18 hours/month × 12 months = as much as 216 IT person/hours saved annually.