Server Consolidation Step 3: Capacity/Demand Management March 11, 2008
Posted by Brad Harris in Data Centers, Disaster Recovery and Business Continuity, Enterprise Architecture, I.T. Optimization, Information Technology, Server Consolidation.Tags: consultant, Enterprise Architecture, Information Technology, ITIL, Server Consolidation
add a comment
Once an accurate inventory of resources and workloads has been taken, the organization must begin to track the utilization of existing resources and the resource consumption of existing workloads. This is necessary so that workloads may accurately be match to target platforms for consolidation.
When tracking utilization, the staff should be sure to monitor for a long enough period to ensure cycles of peak and valley processing periods are included. The monitoring period should include, for example, year-end processing cycles as well as any other business specific events which may have an positive (or negative) effect on resouce requirements.
Once the Inventory and Capacity items have been completed, the organization may begin the analysis required to begin matching workloads to target consolidation platforms. Several factors should be considered. Not only should average workload resource requirements be considered, but prime-shift vs. off-shift changes in requirements should also be considered when choosing workloads for co-hosting. Mixing of production, test and development workloads on the same hardware platform should also be carefuly considered as these often have very different service level requirements. When possible, workloads sharing the same consolidation platform should require similar service, availability and security levels.
At this point, one must also begin to consider the technology being used to facilitate consolidation as this will have some level of influence on the level of resource sharing and consolidation which may be accomplished. This will be discussed with more detail in the next post.
To learn more, visit http://www.innovativetmsolutions.com
Copyright © 2008, Innovative Technology Management Solutions. All rights reserved
Server Consolidation Step 2: Workload Inventory February 19, 2008
Posted by Brad Harris in Data Centers, Enterprise Architecture, I.T. Optimization, Information Technology, Server Consolidation.Tags: consultant, Enterprise Architecture, Information Technology, ITIL, Server Consolidation
add a comment
Step Two : Workload Inventory
The second step in a server consolidation initiative should be an accurate and comprehensive workload inventory. A complete workload inventory should include detailed information on the following attributes :
Applications Hosted - A detailed list of applications hosted on targeted servers must be completed. This list should include any dependencies such as OS verions, hardware requirements, storage, data dependencies, etc.
End Users - The end user base for an application must be understood including the business functions/processes supported.
Service Level Requirements - One must understand in detail the business requirements which drive service levels for the workload. For example, Is the application required to be available 24×7x365 or only 8-5, M-F?
An accurate view of current capacity and demand on that capacity is important in order to make informed decisions regarding consolidation. This information should be used to group various workloads into four categories according to their suitability for consolidation: 1) server co-hosting, 2) application co-hosting, 3) virtual hosting, 4) not a candidate.
Since server consolidation is just like any other architecture initiative, requirements analysis is of the utmost importantce. This work of workload inventory is just that - the gathering of detailed requirements for the final solution.
To learn more, visit http://www.innovativetmsolutions.com
Copyright © 2008, Innovative Technology Management Solutions. All rights reserved
Thoughts on Server Consolidation Methodologies January 30, 2008
Posted by Brad Harris in Server Consolidation.Tags: ITIL, Enterprise Architecture, Server Consolidation
1 comment so far
During my time in I.T., I’ve been fortunate to see many attempts to control server sprawl through server consolidation initiatives. While server consolidation is certainly a worthwile goal, most efforts to reach it ultimately fail for lack of proper planning and failure to take the proper steps in the correct order. Over the next few weeks, I plan to blog here on a recommended sequence of steps for a successful consolidation effort.
Most consolidation initiatives initially focus on hardware (e.g. servers, networks, storage, etc…). While this is certainly a focus that must be addressed at some point, it is a mistake to make this the initial focus of a consolidation effort. The exception to this rule may be when an organization uses hardware footprint reduction as one of the metrics to determine success or failure for a consolidation project.
Step One : Address Organizational Culture
Numerous technologies are available today which facilitate the consolidation of servers. Multiple infrastructure management strategies also exist which can sometimes help reduce server sprawl by leveraging existing resources. Before an organization can realize maximum benefits from these, however, various organizational and administrative issues should be addressed. Once this is completed, one can be successful in consolidating existing workloads as well as driving new implementations toward a consolidated architecture. A clear governance framework should be adopted to ensure that new infrastructure is vetted against the organization’s strategic direction.
Server consolidation should not be considered a project, but a new way of operating a data center. As such, success in this area must begin with a statement of direction by senior management. Policies describing the strategic direction for the organization should be created. These should lay out a direction which includes server consolidation and describes clearly expected benefits and metrics for success. Standard ProcessesStandard processes should be created which implement the goals documented by the governance policies mentioned above. Standard processes will help management to more effectively control technology purchases and decisions and drive them in the strategic direction.
Change Control/Configuration Management
Once standard processes have been implemented, the organization should focus on ensuring that an effective change control and configuration management framework. This is essential so that a process is in place to provide stakeholder visibility into the daily operation of the environment. Change control will further help ensure that modifications to the infrastructure are inline with the organization’s stated goal of server consolidation.
Management Responsibility
Management must “buy in” and accept responsibility for the implementation of the server consolidation plan. Upper management must send the message across the entire organization and publicly/officially support the new direction. This is essential for success. In the coming weeks I’ll post on Step Two : Workload Inventory.
To learn more, visit http://www.innovativetmsolutions.com
Copyright © 2008, Innovative Technology Management Solutions. All rights reserved
I.T. Optimization « Information Technology Insight January 10, 2008
Posted by Brad Harris in I.T. Optimization.Tags: Consulting, Enterprise Architecture, Information Technology, ITIL, Lean Six Sigma, Services
add a comment
I.T Optimization « Information Technology Insight
I.T. Waste
What is it that Security, Change Management, Configuration Management and all of those other quality controls have in common? They’re overhead – waste. It’s true. Think about all of the time, effort and money that is spent in the information technology (I. T.) world in our efforts to make sure things are done right.
What do most security people spend their time doing? They are constantly working to make sure that nothing’s done to open security holes in the I.T. environment. They perform audits, scans, evaluation of new projects and impact changes to the infrastructure. All of this is done in an attempt to keep other I.T. professionals from implementing vulnerabilities. Why? Because the rest of us techies can’t be trusted.
Think about Change Management. Change Management is the process of ensuring that any changes to the infrastructure are properly vetted to ensure that no adverse effects to the environment result. Various specialties are all involved in the process of evaluating these changes : Business Continuity, Security, Configuration Management, Finance, etc… Why? Because bad things could happen if we don’t keep a close watch on everything that goes on.
The same thing is true of Configuration Management and most all of the other “oversight” functions often associated with running an I.T. shop. Why is this important? Let’s consider what Lean Six Sigma (LSS) would have to say about these functions. LSS defines waste as many things… one of which is anything that is required to be done because of a “lack of robustness of a process”. How much of the current work of I.T. would be encompassed by that definition? I would suggest that most I.T. work outside of design, operations and maintenance would be defined as waste by LSS.
So, let’s consider for a moment why (or whether) all of this “wasted effort” is necessary. In general, I.T. managers (and customers) don’t trust I.T. professionals. They don’t trust us to do our jobs in a way that “assures” the confidentiality, integrity and availability of the information that we manage. I think that most other techies out there would probably agree with that statement. Question is: Do we deserve to be trusted? I think that most of the distrust is deserved and is our own fault.
Let’s consider why we as I.T. professionals exist in the first place. Do our employers pay us because we have a particular skill? Or because we’re able to administer the latest cool box out there? Nope… They need us because they build widgets. To build widgets, they need ready access to accurate, and in some cased confidential information. That’s it - period.
The dot-com boom of the late nineties probably did as much as anything to harm the image of I.T. professionals today. Computer geeks were in high demand and took full advantage of the situation. These young companies spent venture capital like there was no tomorrow; much of it on infrastructure. Much of wound upon auction sites after the bubble burst.
Are we really that much different today? For all of our great new technologies and acronnyms, have we really increased the percentage of the I.T. spend that actually contributes to the production of those widgets? In some cases we have due to the adoption of ITIL (and it’s cousins) and Enterprise Architecture. For the most part, however, I think that we still spend far too large a portion of the I.T. budget on stuff that Lean Six Sigma would call “waste”. As I mentioned earlier, those include many things that we all hold near and dear : security, configuration management, audits, management and other quality controls. Again - anything that doesn’t contriute direct and measurable value is realy just waste. That waste may be necessary for some reason, but it’s still waste.
Is it time for I.T. to begin to really focus on the “Leaning of I.T.”? I would suggest that it’s time we begin the development of a new dicipline in I.T. that focuses on reducing the amount of waste in the I.T. budget.
To learn more, visit http://www.innovativetmsolutions.com
Copyright © 2008, Innovative Technology Management Solutions. All rights reserved
