InteQ Corporation
One Van de Graaff Drive
Burlington, MA 01803
Phone: (781) 273-2330
Fax: (781) 273-2215
E-mail: info@inteqnet.com
www.inteqnet.com
InteQ Corporation
InteQ Corporation provides a portfolio of end-to-end service management solutions to its Fortune 1000 clients. These solutions are focused on reducing the total cost of ownership of IT by effective and efficient management of systems, networks, applications, databases, and desktops across the enterprise. InteQ's IT service management solutions unleash the power of IT by integrating people, process and technology to re-focus IT on its customers and align it with the business goals of the enterprise.
TABLE OF CONTENTS
Why Manage your ERP Environment
IT Service Management in the ERP Environment
ITSM Functional Management Areas
Configuration and Change Management
Choosing and Implementing ITSM Tools
IT Service Management (ITSM) is the implementation of a strategy that defines, controls, maintains and enhances IT services for the enterprise. It also bridges the gap between the business community and IT by facilitating communication. ITSM introduces disciplines for supporting the business via Service Agreements -- formal documents that define exactly what is to be delivered and how -- along with the investments required. In order to deliver service level performance and availability, IT must establish a set of service objectives -- from which the objects to monitor and manage and metrics to sustain the environment are derived.
"With a move to distributed computing and client/server applications, the orderly, centralized nature of applications disappeared. The advent of powerful desktop computers and Local Area Networks (LANs) meant that applications could be implemented completely outside the domain of the traditional mainframe environment, and data become decentralized"
[Clowett, Franklin, McCown May 198x). Network Resource Planning for SAPER/3, BAUN IV and Peoplesoft
ERP applications are the consummate example of highly distributed systems. The complexities of ERP implementations pose a challenge to achieving high levels of service for the business community. ERP implementations typically contain complicated database schemas, multiple servers and networks, process hundreds of transactions and have thousands of end user clients. Success of any ERP implementation is highly dependent upon the reliability of the underlying IT infrastructure.
This paper will address the need to monitor your environment, the importance of the planning process and the use of ITSM tools and disciplines to guarantee performance and availability and manage costs. The focus will be on the use of people, processes and technologies particularly in thr areas of:
Why Manage your ERP Environment
Enterprise Resource Planning systems are implemented to run business-critical processes. From order processing through manufacturing, human resources to finance, those systems are the lifeline for enterprises both small and large. ERP implementations performing optimally are key enablers to business productivity and, ultimately, the bottom line. ERR implementations performing marginally can potentially cripple an entire enterprise. Revenue, productivity, and critical decisions can be impaired by ERP system outages. Because of the vital nature of these applications and the significance of the implementation investment, business leaders are expecting commitments to performance and availability from their IT departments.
Additionally, IT departments are evaluated by their ongoing benefit to the business end-users. Unfortunately, end-users and business managers are not always aware of the service accomplishments of the IT organization or the overall health of the application infrastructure – until a prbolem ocurrs. The dynamics of the marketplace and increasing separation between IT and the business has created a chasm between end user perception and reality. In order to close this gap between IT and the end-user, IT must continuously report in the quality of ERP services being delivered as well as the overall business benefits.
figure 3. More complex ERP Infrastructure
Figure 2. Global ERP Implementation
The distributed architecture, which accommodates the highly scalable and flexible implementation of ERP applications, also contributes to the overall complexity of end-to-end ERP system management. Tracking and understanding the source of a problem when one occurs is a daunting task taking days or longer if you have not prepared to manage all aspects of the ERP environment.
Although some ERP vendors have delivered application management tools (SAP’s Computer Center Management System (CCMS)), they fall short of the ability to adequately diagnose problems occurring outside of the application itself. For large implementations with multiple ERP environments (e.g., global corporations that require geography-specific implementations of the ERP application), these tools do not include the ability to consolidate multiple environment events in a single console. This makes management of these applications more cumbersome.
Add to this, the many other objects that comprise your ERP infrastructure (systems, network components, database servers, etc.) and the need to have a consolidated monitoring and management approach becomes apparent. In a research note on ERP high availability, Gartner Group notes that there are third party Network and System Management (NSM) tools that can fill the gap and acknowledges that the configuration of these tools requires "significant effort". But they go on to say that "...despite this, we recommend their use, as most enterprises will experience a short payback period, with the tools often paying for themselves with their first reduction in downtime."
ERP without an IT Service Management Strategy: A Case Study
A 100 Million dollar Internet software company prepared for continued aggressive growth by implementing Oracle Applications ERP system. An accelerated implementation (six months) with heavy business-user involvement during the design phase won kudos for both the CFO and IT Director. The system was comprised of an application server with failover to a second server -- both DecAlpha boxes using Digital TruCluster technology -- and one database server also on an availability cluster. High availability was planned from the onset of the project - from a hardware perspective.
The system came up the first day flawlessly. After several weeks of usage, however, and more users active on the system, performance began to degrade noticeably. Users spanned subnets within the enterprise LAN; however, the users who reported the poor performance seemed to be located within a single subnet. These performance problems were usually reported at the same time of the day. Although the company owned Tivoli enterprise management software, the ERP implementation had been the priority for most of the IT staff so it remained installed but not configured. The systems (database and application servers) were monitored using basic UNIX tools, but the vital signs looked healthy. The LAN was monitored using a sniffer tool, but that looked good as well. The DBA, new to Oracle, checked out the database and deemed that to be performing well. The client workstations were checked to insure that the client hardware and software were not the problem. A week went by and frustrated users continued to see performance degradations. Finally, Oracle was called in to see if they could troubleshoot the problem, which was deemed to be a suboptimal physical data model. This inappropriate configuration should have been easy to detect and solve.
A holistic approach to managing your ERP environment is critical in being able to identify and resolve problems or potential problems. A week of examining every possible object that comprised the ERP environment could have been avoided had proper planning and ITSM disciplines been in place prior to the ERP rollout. The time it takes to understand the cause is directly related to the complexity of the environment. Understanding the role of people, employing effective processes and implementing the right technologies can affect a more proactive management approach to your ERP environment.
You certainly would not consider implementing your ERP system without careful planning and disciplined project management. Likewise, putting a successful ITSM strategy in place requires the same level of planning and project management. The Gartner Group recently reported that 70% of all Enterprise Management packages deployed failed to fully implement. The main reason cited: Failure to plan. Given the rising cost of managing corporate IT and the price tag companies are spending on these applications (approximately 20% of your ERP implementation cost) -- this is a serious matter.
Figure 4. Gartner Group's Assessment of ITSM Tool Implementation Without Planning
Although most IT departments do some form of system, database and network monitoring and management, the type of management and level of sophistication does vary. Your ability to guarantee service levels to your business users will require you to refine your strategy. Regardless of your current ITSM capabilities, it is still imperative to plan for the unique requirements of your ERP processes and infrastructure.
When to Plan
Ideally, we would recommend that ITSM planning be an integral part of the ERP system implementation project plan. Having key pieces of an ITSM strategy in place prior to your ERP system rollout can assist in the following ways:
However, if you did not consider an end-to-end strategy for the management of your ERP infrastructure prior to implementation, it is never too late to begin the process.
IT Service Management addresses the guarantee of Service Levels to the overall enterprise. However, each individual business unit (Human Resource, Finance etc.) will most likely have unique service level requirements. These different requirements must be considered when planning the overall Service Management strategy.
IT Service Management involves careful plans for the use of people, processes and technologies within each of the service management functional areas. The following chart highlights the topics that must be considered when planning your ITSM strategy.
|
People |
Processes |
Technologies |
|
Roles Responsibilities Teams Skills assessment Skills development |
Problem escalation Vendor interaction Change/configuration management Asset and Inventory Call management Communication Service Level Agreements Documentation |
Application Monitoring/Management Database Monitoring/Management Server Monitoring/Management Network Monitoring/Management Desktop Monitoring/Management Help Desk |
IT Service Management in the ERP Environment
As defined in the Overview, IT Service Management is broken down into functional management areas that span information technology segments. The functional areas include -- but are not limited to:
The information technology segments include:
Implementing IT Service Management goes beyond the deployment of technologies to manage and monitor the above functional areas and technology segments. Successful ITSM implementations require that certain levels of sophistication in the management and monitoring of the ERP environment be achieved. Reaching higher levels of sophistication requires enhanced IT management capabilities, process improvements and skill sets to handle the complexities of your business. This is why the planning phase of ERP implementations must include the management of the environment with an end-to-end perspective that includes people, processes and technology.
For instance, in order to achieve an environment for proactively monitoring and reporting on ERP service performance, IT departments must first understand what is in the environment. This requires basic capabilities to discover and poll the objects that comprise an end-to-end ERP infrastructure. One way to accomplish this is through network management applications that present a logical layout of the enterprise.
IT departments can also increase the level of sophistication within the network management technology segment. This can be accomplished by identifying responsibilities for people that manage objects discovered in the ERP environment, defining processes in the event of change or component failure and customizing tools to deliver a business-centric perspective of the infrastructure.
Figure 5. Levels of Sophistication for Performance Management
This figure illustrates an example of the levels of sophistication for Performance Management. Lower levels of sophistication should be in place before higher levels can truly be successful.
In the above example, different levels of monitoring are required for each technology segment. For instance, a server, which is a more critical component in an ERP environment than a desktop client, would require a higher degree of performance management. The level of sophistication required for each IT segment will also depend on your unique business requirements.
Each functional management area has its own unique levels of sophistication as far as people, processes and technologies are concerned -- although there may be overlap. For instance, to efficiently monitor events and faults, IT must first discover the critical ERP objects to manage. Discovering these objects may be handled by the same application used for managing performance of the environment. Regardless, managing events and faults in the environment may require a different level of information from the discovery application than is required for managing performance. In addition, the responsibilities and IT processes for managing events and faults may also be very different from other functional management areas.
The figure below is an example of the levels of sophistication that may be achieved for managing events and faults in the environment. Notice that there are some overlapping functions with Performance Management.
Figure 6. Level of Sophistication for Event and Fault Management
This figure illustrates an example of the levels of sophistication for Event and Fault Management. Lower levels of sophistication should be in place before higher levels can truly be successful.
Achieving ITSM and adhering to service agreements will require different and unique levels of sophistication for each situation.
ITSM Functional Management Areas
Consolidation of events and faults from the ERP environment into a centralized event manager is critical in order to adhere to IT service agreements. This is a major benefit to IT organizations as proactive monitors can be placed within the environment to detect any potential service objective violations. Event & Fault Management requires careful planning to:
Event and fault management differs from Consolidated Service Management in that the event manager accepts faults from proactive management tools that monitor the environment. For example, when an exception or fault occurs in a router, the technology managing that router should be flexible enough to integrate with the event manager by sending a message. Typically, this integration is achieved via SNMP (Simple Network Management Protocol), but may also require proprietary messaging to accomplish this task.
Figure 7. Event Management Flow in an ERP Environment
This example shows the flow of events from various monitoring and management technologies installed within the ERP enterprise. Operations staff interacts with the event manager and utilizes predefined management processes.
Once events are presented to the event manager, IT operators can initiate actions such as escalation procedures or opening a trouble ticket within the service desk/help desk. The event manager is capable of triggering predefined, automated actions that can rectify problems without operator intervention.
Due to the shear number of events that may be triggered, event correlation systems provide IT professionals with root cause analysis that significantly decreases events that are repercussions of the real problem.
Configuration and Change Management
Configuration and Change Management insures continuous operation to the ERP environment by establishing:
Configuration and Change Management disciplines extend to all the components of the ERP end-to-end enterprise. This means that ownership of objects, processes and tools for documenting control and change need to be in place.
Any configuration and change to the ERP environment should go through an approval sequence similar to below:
Figure 8. Change Management Approval Process
The above configuration and change process should be formalized and include a cross-functional team comprised of all organizations impacted.
Performance/Capacity Management
Performance Management presents information to the business about the operation and health of the distributed ERP environment. Business users may require guarantees based on performance needs for the ERP environment such as:
To achieve performance objectives and to make informed optimization decisions, IT must collect metrics from end-to-end ERP components as well as performance perceptions from end-users. IT must:
The ultimate goal is to provide the business with an environment that performs at a level that is consistently acceptable.
Availability Management mandates that ERP applications are available for use during required business hours and avoid incurring downtime costs. Downtime costs include lost revenue, customer dissatisfaction, productivity losses, etc. Ensuring availability in an ERP environment is very difficult to accomplish due to the complexity of its distributed infrastructure. Redundancy and fault tolerance can be built throughout the environment via hardware such as:
However, how are availability service objectives tracked and maintained within IT for business critical applications? Managing the overall availability of critical enterprise applications means:
It is important to note that high availability comes at a price, since redundant technology can be costly.
SAP R/3 for instance inherently has features that utilize server redundancy. However, to protect the network infrastructure that end-users need to connect to SAP R/3 services, external third-party solutions are required. This solution may require multiple network routes to business critical data in the event that the primary network link fails.
Even when fault tolerance or redundancy is achieved, it is still essential to monitor the environment for faults and for fail over occurrences.
Fault Tolerance without Monitoring: A Case Study
A large manufacturing company in the Northeast is utilizing its ERP system for order processing and supply chain management via an extranet to their customers and suppliers. Their Hewlett-Packard server is providing disk mirroring for the database in this three-tier client/server environment. Additionally, the database is utilizing disk striping (tablespaces were stored over several disks) to maximize performance.
One of the disks participating in the database mirroring failed, impacting the majority of the ERP system data. Order processing and portions of the inventory and financial system, were now running on virtually one disk drive. Since IT was not monitoring for this ERP component failure, the event, at first, went unnoticed -- although the disk failure was reported repeatedly to both system and database log files.
Over several days, the help desk staff noticed an increased number of calls from local users and remote customers about the performance of the ERP application, in particular, the Order Processing system. The problem got so bad that the majority of the order processing users could not get the data at all, often receiving strange error messages that indicated a network problem.
The help desk escalated the event as "network problem", but after hours of troubleshooting, the network engineer determined that the problem had to be originating from the database server.
Trying to alleviate the problem quickly, the system manager thought the best and safest course of action was to reboot the system. Unfortunately, the system failed to reboot. After hours of troubleshooting by both corporate and vendor engineers, the cause of the problem was discovered. Once the new disk arrived on site, the system was put back into production.
After rebooting, the system and database log files clearly showed that the source of the problem was disk related. If IT were monitoring these log files, the problem would have appeared immediately after the fault. A new disk could have been "hot swapped" while the system was still running in a matter of hours.
The overall problem resulted in two days of ERP system downtime. The approximate cost of downtime for this manufacturer was determined to be $50,000 per hour.
Consolidated Service Management
The consolidated service desk provides the day-to-day interface between IT and the end-user. Consolidated service management a critical functional area to insure performance, and maintain and report on IT Service Management for an ERP environment.
Figure 9. Consolidated Service Desk
IT staff assigned to the event would perform troubleshooting techniques and may interact with a knowledge base located within the Consolidated Service Desk (see "Consolidated Service Desk). Events are forwarded to the service desk as trouble tickets if escalation is required.
The flow of events from ERP objects into an event and fault management system and then to a consolidated service desk requires careful planning. Defining what constitutes a fault, which events require immediate escalation, etc., must be considered. End-users of the ERP system interact directly with the consolidated service desk for problems, changes, general information, etc.
The Consolidated Service desk includes the traditional help desk functions along with broader support for the ERP infrastructure such as:
End users will require a guaranteed level of support services from IT especially during critical business hours when the ERP infrastructure must be up and running. Organization may require the consolidated service desk to:
The cost of support exists whether or not there is a formal service desk, since end-users will ask co-workers, second level support personnel, etc., for immediate assistance in accessing business critical applications. The consolidated service desk:
Choosing and Implementing ITSM Tools
Implementation Approaches: Building Blocks Vs Big Bang
As stated previously in the section on when to plan your management strategy, it would be ideal to have people, processes and technologies in place prior to going live with your ERP system. But, while it is important to plan the end-to-end manageability of your environment, it may not be practical or prudent to implement the entire architecture at the same time. Whether you choose to do a full-scale implementation of your service management strategy (the "big bang") or implement in phases ("building blocks") depends upon your budget, resource availability and skills, and other projects that are happening simultaneously. It is certainly more desirable to have pieces of the strategy in place working successfully, than to attempt to do everything at once and fail.
In choosing the building block approach, you can determine the priority of the service management components to implement based upon a total cost of ownership analysis of your environment. This study will provide you with data on which management areas will be most effective at reducing the total cost of ownership for your ERP infrastructure based upon your current environment and industry best practices.
Since there are many technology choices available for managing and monitoring the components of your ERP environment, it’s important to understand what to consider when evaluating these tools.
If you are using the building block approach to managing your environment (i.e., implementing Network Management, Database Management, etc. as point solutions to begin to manage components of your environment) keep the following points in mind:
When you are in the midst of planning and deploying your Enterprise Resource Planning application, little else seems to be a priority. However, integral to your ERP project should be consideration of a strategy for IT services in support of your application and infrastructure. The critical nature of these applications to the enterprise has caused business users to expect high levels of performance and availability from IT departments. IT needs to be prepared to meet these expectations.
As with ERP deployments, IT Service Management strategies must be carefully planned. Thoughtful orchestration of the use of people, processes and technologies in support of service objectives is essential. Consider the level of sophistication required to meet set objectives within each management function (event and fault, configuration and change, performance, availability, and consolidated service) and for each of the managed technology segments (applications, databases, networks, and desktops). Remember that you can’t manage what you don’t measure – understand the objects you need to consider in your environment, gather baseline metrics and set thresholds that will alert you to potential problems.
Plan well, but deploy prudently! Be realistic about what you can accomplish as you begin to implement your service management strategy. If you choose to implement in phases, be sure that you have assessed what your deployment priorities should be. Look for ITSM tools that can integrate with your ERP application management capabilities as well as a consolidated event management tool and help desk application to facilitate workflow and enable a coordinated strategy.
Your ultimate goal must to be to insure not just high availability, but high service levels that will maximize the returns expected by the business from their ERP investment.