Technische Universität München boosts data center efficiency with advanced IBM Tivoli Monitoring

Published on 30-Jan-2012

Validated on 18 Jul 2013

"IBM Tivoli Monitoring was our first choice, because it supports our complete cloud-ready hardware and software stack out of the box and allows us to provide the same data to all relevant people, tailored to their individual needs. Additionally, the user-friendly and consistent interface was also a big advantage of this solution over the alternatives we discussed." - Dr. Holger Wittges, SAP UCC Executive Director, TU München

Customer:
Technische Universität München

Industry:
Education

Deployment country:
Germany

Solution:
Automation, BA - Business Intelligence, Information Integration, Small & Medium Business, Systems & Network Management, Workload Optimized Infrastructure Framework, Information Infrastructure, Optimizing IT, Systems Management

IBM Business Partner:
SAP

Overview

This paper describes how Technische Universität München (TU München) deployed a monitoring infrastructure to improve its data center operations for its cloud-ready SAP hosting services. TU München provides over 100 SAP instances for use in courses and training each term to over 160 higher education institutes. Using IBM Tivoli Monitoring, TU München ensures high service levels, enables its support team to provide more accurate and faster feedback to customers, and helps system administrators with management and automation tools to improve their efficiency.

Business need:
The SAP University Competence Center (UCC) at TU München hosts more than 100 SAP systems running in virtual environments on a cloud-ready infrastructure based on IBM Power systems servers, powered by IBM DB2 databases. To increase the reliability of its hosting services, TU München wanted to improve its monitoring solution, to help ensure that it could provide the best service levels to its customers.

Solution:
The TU München team considered three alternatives, and decided to implement IBM Tivoli Monitoring, because it offered the best match with the institution’s complex requirements and provided the best range of features.

Benefits:
TU München has addressed its business challenges with this comprehensive Tivoli solution. Thanks to advanced system monitoring and instant notifications for severe problems, the institution can now identify failures in its cloud-ready infrastructure before customers notice them, helping the SAP UCC at TU München to improve its service levels and increase customer satisfaction.

Case Study

Technische Universität München (TU München) is one of the leading technical universities in Europe, focusing on engineering sciences, natural sciences, life sciences, medicine and economics. The university employs around 7,500 academic and non-academic staff, and is attended by more than 26,000 students.

TU München is member of the global SAP University Alliances Program – a global initiative covering more than 1,100 universities and educational institutions in more than 60 countries, helping to introduce students to modern software applications headed by Elena Maria Ordóñez del Campo, Senior Vice President, SAP AG. As one of two SAP University Competence Centers (UCC) in Europe, TU München provides educational services, hosting, backup and recovery services for over 100 SAP instances, somewhat comparable in character and size to real-world production systems, used by over 160 affiliated institutions of higher education.

To support the university, SAP donates the latest software releases such as SAP Business Suite 7, SAP BusinessObjects solutions and SAP NetWeaver technology and components and organizes free training courses. TU München maintains close relationships with the International SAP IBM Competence Center in Walldorf and the IBM Solution Sales for SAP team within IBM Germany, which both provide support for TU München.

IBM supports TU München’s SAP projects by providing the IT infrastructure, based on the IBM Power Systems (POWER7) platform and XIV storage, including operating systems, database and virtualization technology, managed with the IBM Tivoli suite of applications. IBM offers intensive knowledge transfer and ongoing support through the regular presence of highly-qualified IT architects and IT specialists on site.

Based on this support, TU München provides the SAP systems supporting its education courses and research projects at relatively low costs, on an ultra-modern systems landscape.

Initial IT environment

Before the implementation of IBM Tivoli Monitoring, TU München used SAP Solution Manager to monitor its SAP systems. SAP Solution Manager provides functionality that covers solution deployment, operation, and continuous improvement. This centralized, robust application management and administration solution combines tools, content, and direct access to SAP to increase system reliability.

While the SAP Solution Manager is appropriate for monitoring SAP applications, the team could not monitor the entire underlying hardware and system software solutions end-to-end.

To complement the SAP application monitoring, TU München’s IT team developed some reports to cover additional aspects such as database backups and selected hardware components with its monitoring. The extra reports were cluttered and not accessible from a shared portal. The only way to find and access the reports was by knowing the usually undocumented locations of the reports on internal web servers. Moreover, the reports were difficult to decipher, as they contained no detailed error reporting beyond the simple plain numbers. To benefit from the report at all, substantial background knowledge was needed to interpret the reported information correctly.

With more scripting, the in-house solution could have been extended to cover more monitoring aspects. This alone would not address the ‘findability,’ usability and information content of the reports. The TU München IT team chose to replace the existing solution, and looked for an easy-to-use, well-documented and easily extensible standard solution.

Business challenges and project objectives

The SAP UCC at TU München hosts more than 100 SAP systems running in virtual environments on a cloud-ready infrastructure based on IBM Power systems servers, powered by IBM DB2 databases. To increase the reliability of its hosting services, TU München wanted to improve its monitoring solution, to help ensure that it could provide the best service levels to its customers.

In the past, some SAP system outages were detected and reported by customers before the IT team was aware of an issue. The TU München IT team wanted to get back into the driving seat, so to speak, by introducing comprehensive and integrated monitoring that would help to identify, rectify and prevent outages. Any new solution would also replace custom-developed reports, standardizing the monitoring system and making it easier to maintain and extend.

The legacy solution covered relatively few aspects of the overall IT environment, as it lacked sophisticated support for monitoring server hardware components, could not monitor devices such as routers, and was unable to include software components such as database systems. It presented a complex user interface that was difficult for infrequent users to remember and understand, which restricted the number of people who could access and exploit the reports.

“We expect a significantly improved system monitoring environment as a result of this cooperation with IBM,” says Professor Helmut Krcmar, SAP UCC Academic Director. “We want to replace a number of isolated solutions with a comprehensive monitoring infrastructure. With a complete stack of IBM hardware and software, we are confident that the implementation of IBM Tivoli Monitoring will boost the availability and reliability of our SAP hosting services.”

The main objective for TU München was to establish a central, comprehensive monitoring solution to ensure reliable SAP operations for its customers. This would address the shortcomings of the legacy solution by including both the hardware and the software stack. The institution wanted to monitor more than one hundred LPARs, most of them running SAP systems, and some providing infrastructure services to the other systems. The selected monitoring solution should be able to support yearly growth of 20 percent in the number of LPARs. Additionally, it should also include every component contained in TU München’s cloud-ready infrastructure, from networking equipment, routers, server hardware, virtualization layers, to the operating system, database management systems and SAP applications. The objective was that all of these would be continuously and automatically monitored.

Additionally, the TU München team wanted to take into account the needs of the four main user groups, who have different needs in terms of access to status reports and advanced analytics tools, based on the monitoring data. First, there are the system administrators, who need to work with monitoring data on a daily basis. Second, support staff should also be able to check system status information and drill-down into issues to provide reliable information to people reporting problems. Third, managers should be able to discover information about their team’s systems. Fourth, selected status and performance data should be made available to customers.

As TU München is an educational institution, another objective was to implement a solution that is widely used in the commercial world, so that experience gained on the system would be beneficial to team members such as scientific staff and student assistants in future.

Finally, TU München wanted to introduce a solution that supports new SAP applications quickly, without the need to develop one-off adapters and clients. This required choosing a solution with comprehensive vendor support, and one which has an ongoing, active development strategy, able to take advantage of future technological improvements.

Selecting IBM Tivoli Monitoring

The TU München team considered three alternatives, and decided to implement IBM Tivoli Monitoring, because it offered the best match with the institution’s complex requirements and provided the best range of features.

The principal reasons for selecting Tivoli Monitoring were the seamless integration with the cloud-ready IT landscape, based on IBM hardware and software, and the advanced integration capabilities with other systems. The Tivoli package includes a huge number of pre-built device adapters, and the flexible system architecture enables new adapters and components to be added rapidly and easily.

Dr. Holger Wittges, UCC Executive Director at TU München, “IBM Tivoli Monitoring was our first choice, because it supports our complete cloud-ready hardware and software stack out of the box and allows us to provide the same data to all relevant people, tailored to their individual needs. Additionally, the user-friendly and consistent interface was also a big advantage of this solution over the alternatives we discussed.”

Technical implementation and architecture

To allow universities and other educational institutions to teach students concepts and approaches of modern business applications, TU München provides a complete range of SAP solutions to its customers. The solutions include SAP Business Suite 7, SAP NetWeaver Business Warehouse 7, and SAP BusinessObjects applications.

TU München operates a fully cloud-enabled infrastructure based on the IBM Power Systems platform, which automates many administration tasks and greatly accelerates preparation for the new semester.

Working with IBM, TU München implemented IBM Tivoli Monitoring in a four-tier architecture that is capable of managing and monitoring this virtualized hardware and software landscape:

  • Monitoring agents collect raw data from different hardware platforms, database systems and applications.
  • Tivoli Monitoring servers handle data management tasks and control the agents.
  • Portal servers provide access to user-friendly visual presentations of the data.
  • IBM Tivoli Enterprise Portal clients.

IBM installed and configured the core components within six days, and the entire implementation was completed in less than one month. After the initial setup, TU München staff spent five days on defining alerts and notification options. Subsequently, supported by IBM, the team worked on automating the deployment of the solution and scripts to generate customized standard views. After this work was completed, IBM trained the TU München team and answered questions that emerged during the implementation. Finally, after a test phase, TU München rolled out all the clients and the new monitoring solution was deployed.

The team installed IBM Tivoli Monitoring 6.2 on three LPARs running the AIX 6.1 operating system. The LPARs are set up on an IBM Power 750 server, which is connected to an IBM XIV Storage System. The monitoring landscape is configured and controlled centrally, and TU München and IBM set up one Tivoli Enterprise Monitoring Server as the “master.” This master server connects to multiple remote monitoring servers, which collect data from the LPARs, underlying hardware systems and networking equipment. The master server receives and processes the data, and forwards it to an IBM Tivoli Data Warehouse.

An IBM Tivoli Enterprise Portal server is directly connected to the master server, providing centralized access to up-to-date monitoring data and system status information via Java-based and browser-based graphical user interfaces.

For advanced analysis, an IBM Cognos Business Intelligence server will be connected to the IBM Tivoli Data Warehouse. To provide access to the data with a consistent look and feel, the Cognos Business Intelligence server will be connected to an IBM Tivoli Integrated Portal Server that will deliver sophisticated reports through a modern web interface.

This advanced system architecture, which features load balancing for the remote monitoring servers, enables TU München to monitor its large SAP landscape easily.

TU München needs to make sure that resource assignment and allocation for over 160 customers are operational. IBM Tivoli Monitoring supports TU München in that task by collecting data points from more than 800 agents deployed across the cloud-ready SAP environment.

This environment is constantly growing in size and complexity. The agents include features to monitor SAP systems, IBM DB2 database management systems, AIX operating system functions, UNIX log files, IBM Virtual I/O Server instances, IBM PowerVM virtualization, IBM Hypervisor, IBM Hardware Management Console equipment and also the core of the POWER7 servers, the IBM Central Electronics Complex (CEC) modules.

To roll out new systems quickly, TU München automated the deployment of the agents monitoring IBM AIX operating system functionality, IBM DB2 functionality, and the SAP applications. These agents are now integrated into the general automated LPAR provisioning solution at the SAP UCC.

For the monitoring agents, the team discussed, developed and tested configurations, aimed at reducing the total amount of monitoring data and allowing TU München’s system administration staff to monitor the large amount of systems and components as efficient as possible. For each agent, the team used predefined alerts and configured additional individual alerts, to make sure that failures will be prompted and reported quickly. This helps to ensure that the operations team is able to resolve issues before customers notice any problems with their applications.

Alerts are displayed in the Tivoli management interface, and can also trigger emails and text messages, and further channels can be used to notify support staff as needed. Furthermore, alerts can trigger the execution of scripts, to automate more complex notifications and actions. At TU München, most alerts simply pop up as notifications in the management interface. The most important alerts also trigger emails. TU München staff members carry up-to-date smart phones, which allow them to receive email alerts instantly, regardless of the time of day.

After examining the comprehensive capabilities of the monitoring solution, the team decided to select the indicators most important to TU München. The team installed and configured agents and alerts to monitor all aspects of the SAP environment.

At the hardware level, IBM Hardware Management Console and IBM Central Electronics Complex agents monitor the physical state of the server to detect hardware failures as early as possible. At the hypervisor level, the IBM Tivoli Monitoring solution uses IBM Hardware Management Console and IBM Central Electronics Complex agents to monitor the availability of processors and main memory to alert if resource and capacity issues could be identified.

At the operating system level, the team deployed Unix and IBM Premium AIX agents to monitor each system’s file system size, the existence and the size of a list of specified files and folders, as well as the existence of important system processes. This ensures that the team gets immediate notification when crucial operating system resources or configurations are not operational. For example, the system monitors the availability of the SAP router system processes as well as the status of the file volumes, to make sure that the systems do not run out of storage space.

To monitor the database system, the IBM DB2 monitoring agent is installed, configured to monitor the last backup date as well as if the current IBM DB2 instance is running and if the status of the relevant database is active. This configuration ensures that the required database resources and backups are available.

At the application level, the SAP UCC configured the SAP monitoring agent to check the availability and performance of the corresponding SAP instance and monitor the system configuration. Monitoring these SAP system characteristics enables the SAP UCC proactively to alert the IT team as soon as any SAP systems encounter failures.

All monitoring, including the IBM XIV Storage System with its integrated monitoring component are configured to provide proactive monitoring. The monitoring system automatically sends notifications to relevant staff, if failures are detected.

Integrated and consistent user interface to access status information

Access to monitoring data is provided through Java-based and browser-based applications. Data collected is represented within one single user-friendly point-and-click interface. The user experience is supplemented with a number of standard views, alongside customized views based on the advanced configuration options that the solution provides.

In extending the standard views, TU München enhanced the data pool with information about what systems belong to which customer, which enables customer-oriented access to monitoring data. Customer data is imported from an existing system support and billing database that the University maintains. If a customer calls with an issue, the customer-based view enables support staff and others to drill down into the customer’s system data to locate and identify problems without knowing in advance which resource, application, service or device might be the cause.

The standard view on the monitoring data provided by the IBM Tivoli Monitoring solution is a resource-oriented view. Instead of customers, a single resource is selected and then based on this one can drill down into the data starting from the specified resource.

Important to TU München was to provide a central and consistent user interface for different users. TU München identified four categories with different requirements: administrators, support staff, managers and customers.

Administrators

Administrators are the people who rely heavily on the solution and use it frequently. For administrators, the system must provide a clear and quick overview, notify them if issues arise, and provide them with tools to monitor and manage the cloud-ready infrastructure efficiently. Administrators can customize and extend their reporting views easily, without the need to manually write and edit custom-developed monitoring scripts in various locations.

Administrators can also trigger mass operations directly from within the monitoring application. This can be useful, for example if a group of systems needs to be restarted. With IBM Tivoli Monitoring, the administrator specifies the command and selects the system where the command should be executed. The mass-command complexity, such as login to the local systems and actually executing the command, are handled by the monitoring system.

Alexandru Danciu, Research Associate, says, “In the past, executing mass operations on a couple of systems was a pain. The feature to trigger mass operations from within the monitoring and systems management interface made system maintenance and trouble shooting a lot easier.”

Administrators also benefit from the integrated interface with drill-down capability to locate errors quickly, no matter where, exactly, the failure occurred.

Support staff

To provide efficient help and assistance, TU München’s support team needs to be able to access monitoring data. The institution implemented different views, to provide the support team with different ways to access monitoring data depending on the support request.

For example, support staff can drill down into the data based on systems grouped by customers, by SAP solutions and by server systems. With the new monitoring solution, support staff can quickly connect to different systems and check the status of components such as the database, right at the time when a customer is calling. Providing this detailed view to the support teams allows the staff to be better informed with high-quality, accurate data at their fingertips.

Allowing support staff to access the monitoring data also improves the quality of the support requests sent to system administrators. Responses are much more detailed, because the cause of a problem has already been identified by the support team. Administrators are freed from checking multiple systems to check if they are the problem source, and instead can turn immediately to resolution.

Managers

With the IBM Tivoli Monitoring solution, TU München can provide managers with a clean and simple user interface, which looks similar to modern desktop applications. Occasional users who are familiar with standard presentations and spreadsheet applications find it easy to work with the monitoring data, and can view information by customer, products or resource. Using the IBM Tivoli Data Warehouse together with IBM Cognos Business Intelligence technology, managers will also be able to view data as charts, and visualize and analyze monitoring data over time. This visual trend analysis will enable managers to identify irregularities, and report on key performance indicators.

In future, TU München also wants to provide customized dashboard views to managers, to provide them with aggregated systems status overviews.

Customers

TU München plans to provide its customers direct access to their systems status information. The goal is to include information about the number of systems they currently have, how many of them are running, how many users are logged in, how much data the systems are using and many other key performance measures.

The next step will be to implement self-service features based on IBM Tivoli Service Automation Manager, to enable customers to order SAP systems online. Once this is implemented, TU München is planning to enable customers to view and analyze relevant monitoring data and to trigger backups and related actions for themselves.

Tivoli Data Warehouse and IBM Cognos Business Intelligence

A key component of the monitoring solution that will be implemented at TU München is the IBM Tivoli Data Warehouse, a data warehousing solution based on proven IBM DB2 Data Warehouse technology. The system enables TU München to analyze and process historical systems management data to compile performance reports, gain new insights and predict possible future issues.

A data warehouse is a comprehensive pool of data, collected and stored, that can be analyzed and visualized along different dimensions and processed using complex algorithms. The purpose is to provide access to large quantities of data and explore, analyze, and understand the information.

The IBM Tivoli Data Warehouse is an online analytical processing (OLAP) system, designed to store and manage historical data from both Tivoli and other management applications. IBM Tivoli Data Warehouse aggregates and correlates the data from multiple systems for use in reports, and in analytics solutions.

To provide managers with accurate reports and allow them to analyze the performance and status of their systems over time, TU München will implement IBM Cognos Business Intelligence and connect it to its monitoring data warehouse. With IBM Cognos Business Intelligence at hand, managers will be able to dive into the data and examine it to gain new insights and to support decision-making processes.

Furthermore, to optimize systems operations and business continuity, IBM Tivoli Data Warehouse will be used in combination with the IBM Tivoli Performance Analyzer and IBM Cognos Business Intelligence to predict future trends and behaviors, enabling TU München to make proactive, knowledge-driven decisions with the help of advanced data mining tools.

TU München generates reports such as CPU, memory, disk and network utilization in different visualizations, such as a heat chart. When the implementation is complete, these reports can be accessed through IBM Tivoli Integrated Portal, an integration solution that ensures that the user interface of the main IBM Tivoli GUI and the reporting and analytics solution are consistent, to keep the systems as easy-to-use as possible.

Project achievements

TU München has addressed its business challenges with this comprehensive Tivoli solution. Thanks to advanced system monitoring and instant notifications for severe problems, the institution can now identify failures in its cloud-ready infrastructure before customers notice them, helping the SAP UCC at TU München to improve its service levels and increase customer satisfaction.

With IBM Tivoli Monitoring’s automation capabilities, the TU München team has also automated reactions for recurring problems. This reduces the workload of the IT team members, and allows them to concentrate on more complex issues. The Tivoli monitoring infrastructure enables the UCC to run its SAP environment with a comparatively small IT team. Supported by the IBM Tivoli Monitoring solution, only one person is required to ensure continuous system operation. If a problem is detected, this person gets notified instantly. To solve issues when they arise, the entire IT team of ten people works together to provide continuous availability of the hosted SAP systems.

By integrating the installation of the IBM Tivoli Monitoring agents into the automated Cloud LPAR provisioning solution, TU München saves huge amount of time when deploying new SAP systems. Previously, it took two hours to manually install and configure the different agents; now agent deployment can be completed in one minute at the push of a button.

The new IBM Tivoli Monitoring solution also boosts the efficiency of support staff. When customers call to report that they cannot access their system, the support team can now check directly, and within two minutes, in the user-friendly IBM Tivoli Monitoring interface. Support staff can see where the monitoring solution system noticed a failure, and which other systems are running without problems. In the past, the support team could not find this out themselves, or it could take about half an hour to validate the customer’s report.

“The new monitoring infrastructure has improved the performance of our support team substantially,” says Holger Wittges. “Typical support cases can now be verified five times faster than before, and the information provided to the customer is much more accurate.”

TU München plans to implement IBM Tivoli Data Warehouse software, combined with IBM Cognos Business Intelligence technology for advanced analytics at a later date. This will enable the university to benefit from much deeper insights into system utilization and the overall cloud-ready data center infrastructure. Providing detailed performance reports and an easy-to-use ad-hoc reporting interface to managers will enable them to use system monitoring data to support decision-making processes.

“Improved reporting and more accurate information improves decision-making a lot,” says Helmut Krcmar, SAP UCC Academic Director. “The next step in using business analytics tools will be to enable administrators to use the data warehouse for advanced data mining and predictive analysis, to allow them to develop a proactive approach to systems management.”

Next steps

Future plans at TU München include the extension of its IBM Tivoli Monitoring solution to provide customer access to system status information. The overall goal is to expand the existing IBM Tivoli Service Automation Manager solution, implementing a fully automated SAP hosting service, including customer self-service features to set up, manage, and monitor complete SAP landscapes.

The cloud-ready IBM infrastructure was the first step, followed by basic automation of Cloud LPAR provisioning. By adding a consistent system and application monitoring solution, TU München has created the foundation for further optimization and automation.

To raise the level of automation, TU München and IBM plan to implement more advanced event management with IBM Tivoli OMNIbus. This solution will provide the features to enable TU München to calculate complex correlations between different events and alerts.

The correlations will then be complemented by sophisticated rule definitions to allow TU München to automatically react accurately to complex situations in its IT systems. All these features will be provided with a consistent user interface through IBM Tivoli Integrated Portal server.

TU München also plans to extend the monitoring solution itself, by integrating the latest version of the SAP Solution Manager into the existing IBM Tivoli Monitoring infrastructure.

To monitor actual system performance from an end-user’s point of view and establish end-to-end performance analytics of the cloud-ready SAP infrastructure, the team is also considering the implementation of additional external monitoring agents to check system availability based on GUI access to the SAP systems.

Products and services used

IBM products and services that were used in this case study.

Hardware:
Power 750, Power Systems running AIX 6, Storage, Storage: XIV

Software:
Tivoli Performance Analyzer, Tivoli Service Automation Manager, Tivoli Storage Manager, Tivoli Monitoring, AIX, DB2 for AIX, Cognos Business Intelligence

Service:
IBM-SAP Alliance

Legal Information

© Copyright IBM Corp. 2012. IBM Deutschland GmbH, D-71139 Ehningen ibm.com. Produced in Germany, January 2012. IBM, the IBM logo, ibm.com, i5/OS, Cognos, DB2, Domino, FlashCopy, Lotus, Notes, POWER, POWER4, POWER5, POWER6, System i, System x, and Tivoli are trademarks of International Business Machines Corporation in the United States, other countries, or both. If these and other IBM trademarked terms are marked on their first occurrence in this infoarmation with a trademark symbol (® or ™), these symbols indicate U.S. registered or common law trademarks owned by IBM at the time this information was published. Such trademarks may also be registered or common law trademarks in other countries. A current list of other IBM trademarks is available on the Web at: http://www.ibm.com/legal/copytrade.shtml UNIX is a registered trademark of The Open Group in the United States and other countries. Linux is a trademark of Linus Torvalds in the United States, other countries, or both. Microsoft, Windows, Windows NT, and the Windows logo are trademarks of Microsoft Corporation in the United States, other countries, or both. Other company, product or service names may be trademarks, or service marks of others. This brochure illustrates how IBM customers may be using IBM and/or IBM Business Partner technologies/services. Many factors have contributed to the results and benefits described. IBM does not guarantee comparable results. All information contained herein was provided by the featured customer/s and/or IBM Business Partner/s. IBM does not attest to its accuracy. All customer examples cited represent how some customers have used IBM products and the results they may have achieved. Actual environmental costs and performance characteristics will vary depending on individual customer configurations and conditions. This publication is for general guidance only. Photographs may show design models.