Modeling Assumptions

The X_factor modeling algorithims are based upon the following assumptions. These assumptions qualify the solution to handle most situations optimally. Any specific requirements in terms of business, technical, organizational, or procedural policies, requirements, and limitations will require further analysis to determine the need for additional resources to build a suitable virtual infrastructure solution.

Target Platform

• All target virtualization host servers will be identical in configuration.
• All NIC ports in the target hosts will be active. If not all ports will be active, more hosts may be required to distribute the potential network load to avoid network-based performance bottlenecks.
• The use of Gigabit NICs assumes a network using Gigabit speeds.
• Shared storage using a SAN will be available.
• All target server hosts will have access to required storage, and storage limitations are not considered a gating factor.
• An additional server and database will be required for the management server and system.
• All workloads will utilize virtual disks residing on a virtualization file system, and no virtual disks will map directly to SAN LUNs.

Current Utilization

• The sampled servers are representative, and any extrapolations to be made of their utilization metrics to a larger server universe must assume no dramatic differences in utilization.
• The current utilization rates observed will not increase significantly in the future.

Future Utilization

The target hosts are reserved exclusively for consolidation of the existing servers and server containment to support future provisioning requests. These hosts do not account for unplanned workloads other than for a contingency or workload balancing.

All target hosts will be utilized to their maximum potential and workloads may be balanced across the target hosts with no network limitiations.

• The scenario given is based on information received in the collected workload data. It is meant to serve as a real example on how the implementation could be built. Other scenarios using alternative configurations can be produced in addition to the included example.

Candidate Inventory

The uploaded data collection is used by X_Factor to create a list of Candidates. This list is essentially an inventory of each host and allows you to review the performance, OS, CPU, RAM, machine type, etc. for each. Aside from having an inventory of your environment, this list will also be helpful if you are considering multiple modeling scenarios.

Candidate Inventory Review

• To review the candidate data click on "Data Collection" on the left side of your screen and then click on "Candidates" from the drop down menu.
  
• A list of all the candidates and their collected data is provided as shown below. Please look over the list to familiarize yourself with its contents. Detailed capacity and performance data on each host is available by clicking on the "Details" button.
  
  Detailed capacity and performance data is shown below. You will notice there is a "Score" box for Total, CPU, Memory, etc. This box will be empty until you complete the modeling task. Note: A scroll bar is located on the right side of this screen. Be sure and scroll down to see all the elements for this host.
  

Sorting Options

The column headings may be used to create different sort options. The default display of the inventory is alphabetical (A-Z) by the "Server" name as shown below. You can change this sort to (Z-A) by clicking on "Server". The two views are shown below.

Note: The small black arrow within the selected column heading points up or down to show the sort order.

(A-Z)

(Z-A)

Rules for Sorting

You can achieve different sorts by clicking on the different column headings. The default is in ascending order whether it is a numerical or text column. For columns that contain both text and numbers, such as, "OS", the list will be sorted alphabetically and then numerically. The sort options here are not used by other X-Factor modules, rather they provide different views of the inventory that can be exported to Excel spreadsheets.

These inventory lists usually contain multiple pages depending on the number of hosts.

• Click on the arrow on the bottom left of each page to advance or decrement the page being viewed. You may also modify the number of rows per page if desired.
  
• Also on the bottom of each page is an "Export" button. You may export this list in the form of an Excel spreadsheet and save it to your local PC. To do this click "Export" and complete the "File Download" prompt. Note: Clicking Export will download the entire Candidate file, not just the page being viewed.

Below is a partial example of an exported Candidate list.

Server	CPU Core Count	CPU Speed (MHz)	RAM (MB)	OS	OS Version	Chassis Make	Chassis Model
CMS001	1	299	512	Microsoft Windows 2000 Advanced Server	5.0.2195	None	None
DEM-WEBEOC1	4	2405	8023	Microsoft Windows Server 2003, Enterprise Edition	5.2.3790	HP	ProLiant DL385 G1
DEM-WEBEOC3	8	2833	3325	Microsoft Windows Server 2008 Enterprise	6.0.6001	HP	ProLiant DL380 G5
FIRE-APPS1	4	3200	3071	Microsoft Windows Server 2003, Enterprise Edition	5.2.3790	HP	ProLiant DL380 G4
FIRE-APPS10	8	1999	24573	Microsoft Windows Server 2008 Enterprise	6.0.6001	HP	ProLiant DL380 G5
FIRE-APPS10V	1	3000	8191	Microsoft Windows Server 2008 Enterprise	6.0.6001	VMware Virtual Machine	VMware Virtual Platform
FIRE-APPS11	8	3000	3325	Microsoft Windows Server 2008 Enterprise	6.0.6002	HP	ProLiant DL380 G5
FIRE-APPS12	8	1999	3325	Microsoft Windows Server 2008 Enterprise	6.0.6001	HP	ProLiant DL380 G5
FIRE-APPS13	2	1800	2047	Microsoft Windows Server 2003, Standard Edition	5.2.3790	HP	ProLiant DL385 G2
FIRE-APPS14	2	2664	2030	Microsoft Windows Server 2003, Standard Edition	5.2.3790	Intel	DP965LT_
FIRE-APPS15	2	2612	2047	Microsoft Windows Server 2003	5.2.3790	None	None
FIRE-APPS16	2	2612	2046	Microsoft Windows Server 2003, Standard Edition	5.2.3790	Sager Notebook Computer	S2865
FIRE-APPS17	2	2787	3071	Microsoft Windows Server 2003 R2, Enterprise Edition	5.2.3790	HP	ProLiant DL380 G3
FIRE-APPS18	2	3049	5631	Microsoft Windows Server 2003 R2, Enterprise Edition	5.2.3790	HP	ProLiant DL380 G3
FIRE-APPS19	2	3400	2047	Microsoft Windows Server 2003 R2, Enterprise Edition	5.2.3790	HP	ProLiant DL380 G4

Select Workloads

A typical IT environment may contain a variety of physical hosts, virtual hosts, applications, service levels, locations, domains, funding centers, etc. Select Workloads is a task that enables you to choose which hosts you will model as part of your virtualization planning. You may model all hosts together or separate them based on your business requirements. If you choose to model the hosts in separate scenarios, we recommend that you do so by creating separate projects. Note: Since the uploaded data collection is tied to a specific project, you will need to upload the file into each project you create.

Topics covered here include:

• Including Workloads
• Including All Workloads
• Excluding Workloads
• Including a Subset of Workloads

Including Workloads

From the Project Status screen click on "Data Collection" and then click on "Select Workloads" from the drop down menu, as shown below.

The "Workload Selection" screen below will be displayed. This screen displays the same list of the candidates that were shown on the Candidate Inventory list only in a different format. Note: When this list is first viewed the default "Status" shows that all workloads (hosts) are excluded from being sent to modeler.

This screen allows you select which workloads (hosts) you want to "include" for modeling purposes. You may include all the workloads or you may choose only a subset for modeling.

Looking over the contents of this page you will see a series of column headings (e.g. Workload from Inventory). Below each heading you see an empty box. These boxes can be used to refine the sorting and selection process. On the left hand side you see a series of check boxes that are used to select specific hosts on the list. Below the list you see page numbers and a button called "Create Filter". Along the bottom of the page you see a series of buttons that aid in selecting hosts and in sorting the list. Each of these areas will be discussed in more detail as they apply during the following discussion.

Including All Workloads

1. If you decide to include all the hosts in the modeling phase simply click on the "Select All" button on the bottom of the page. The Select All button will put a check mark in each box to the left of every host name on every page and highlight every row as shown below.
   
2. Next you click on "Change Selected" and a dialogue box will appear asking what you want to change the status to (Included or Excluded). Below that prompt is a space for describing the reason for a particular status. Usually this space is used to explain Exclusions and any entered reason entered here will be displayed in later document generation. The red arrow shows the "Change Selected" button (see below)
   
3. To complete the status change click "OK".

   As shown below you will see that the status for each Host has been changed to "Included".
   

   The Project Status task list will be updated to reflect that you have selected workloads.

   You can navigate back to the Project Status task list from any page within X_Factor by doing the following:

4. Click on the "Favorites" tab on the left side of the page and then select "Client Summary" from the drop down menu. Please note that "Select Workloads" is now highlighted in green and has a check mark next to it.
   

Excluding Workloads

You may want to exclude some hosts from modeling for business reasons like a specific application requires its own physical server.

1. To do this, go thru the "Workload Selection" list and check the box to the left of each Workload (host) name that you would like to exclude. In the example below, I have checked two hosts.
   
2. Next you click on "Change Selected", then change the status in the dialogue box to "Excluded" and enter the reason if desired as shown below. Any entered reason will be displayed in later document generation. The red arrows indicates location of "Change Selected" button.
   
3. To complete the status change, click on "OK".

   The screen below now shows the status changed to "Excluded".
   

Selecting a Subset of the Workloads for Modeling

It is not uncommon for a client to model specific groups of hosts separately. The business reasons behind this could be related to the application, machine type, operating system, location, funding, already virtualized, etc. Plus, some people just want to model different scenarios.

To include only specific hosts for the modeling phase you can select those hosts individually from the entire list or, if they have something in common like the host name followed by an incrementing number, you can sort the list using a specified criteria and create a new list that only includes the host names which match that criteria. This new list can be created by entering criteria in just one heading or you can refine the list by entering criteria in multiple headings. In this case, he default rule will use "AND" when criteria in multiple headings is used to determine the new list. You can then sort the new list by other variables which will also be covered in this topic.

For demonstration purposes we will be using a data collection consisting of 67 workloads (hosts).

The first step is to ensure the status for all workloads is "Excluded" as shown below. To change the status to "Excluded", click on "Select All", then click on "Change Selected", select "Excluded" from the popup menu and then click "OK".

Please note that the list below is Page 1 of 5 Pages that contain 67 items and is shown in ascending alphabetical order.

It is important for you to understand the location and purpose of the various fields and buttons that can be used to aid in selection of workloads. On the screen below you will see red arrows that point to different fields and buttons. These are described below.

1. This area consists of a series of column headings contained in a light blue background. By clicking on a particular column heading the list will be sorted alphabetically by the heading in that column. For example, if you click on the "Workload From Inventory" heading the list will be sorted alphabetically by host name. The little triangle on the right of the heading (directly under red arrow) indicates whether the sort is in ascending or descending order. Only one heading may be selected at time. The little triangle will popup within the heading that has been selected.
2. Under each column heading is a blank search field. By entering alpha and/or numeric's this field can be used to create a refined list based upon your criteria. If you enter criteria in more than one field they will be "ANDed" together. To the right of the search field is a "key-like" symbol. When clicked you will see a series of search rules that apply to that column. The default for each column is "Begins with". Examples to follow.
   
3. The "Create Filter" button allows you to apply a set of Boolean rules to your criteria. The default is "AND".
4. The "Field Chooser" button allows you to sort your new list by various categories.

Sorting by Column Heading

Clicking on any column heading will enable an alphabetical sort function. On the screens below notice the little triangle in the "Workload From Inventory" heading. That indicates two things. First, it means that you have clicked on the heading to get a sort by Host name. Secondly, the pointing direction of the triangle indicates the sort is done by ascending or descending order. The first screen shows the arrow pointing up. The second screen shows the arrow pointing down. Note the difference in the lists displayed.

The screen below shows a sort by "Make".

Creating New Lists using the Blank Search Field

These fields allow you to enter alpha/numeric criteria to locate and list specific workloads. You can enter a single alpha or numeric or a string depending on how specific you want the criteria to be. The "key-like" symbol on the right side of each field displays a list of qualifiers that tell X_Factor how to use the alpha/numeric string. For example, if you wanted to model only the Exchange servers from the list below you could set the qualifier as "Begins With" and enter search criteria of "fire-e". The more characters you enter into the field the more refined your search will be. Note below that X-Factor located three (3) servers matching the criteria of "fire-e", all of which are exchange servers.

Note: To model the new list you must redo the candidate selection process and change their status to "Included"
Each time you create and model a new list, the modeling data for the previous list is over written by the new modeling data. You can do two things to permanently retain all your modeling data.

1. You can create separate projects for each modeling scenario.
2. You can complete all the steps on your project task list which would give you Word documents containing your modeling data and results.

The search field is available within each column that allow you to create a list based upon an individual column or use multiple columns for refinement. When you use multiple columns to search, you need to set the "qualifier" (key-like symbol) within each column. The default for all columns is "Begins With". In addition, the default Boolean rule is "AND" For example, the workload list below was created by searching for all FIRE-APPS that are running on a Dell server. The search criteria used was "fire-a" within the Workload From Inventory column, "d" within the Make column, a qualifier of "Begins With" and a Boolean rule of "AND". The bottom arrow is pointing at the create filter field. Note that your search criteria is shown here. You may have to zoom your screen to see it.

Since only one drop down menu is displayed at one time, the screen below shows the dropdown menu for Boolean rules.

To see which Boolean rule is selected click on the Create Filter field (red arrow below). You will see a popup called "Filter Builder" that shows you the rule in RED and your column search criteria. In this case you see the default rule of AND.

To change the rule, click on the one shown in red and another pop up list will be displayed. Select the rule you want to apply and click OK. See below.

Field Chooser Button

On the bottom, right of the screen is the "Field Chooser" button. This button allows you to sort your new list by server characteristics not visible on the Workload Selection screen. When you click this button a popup menu appears that displays these characteristics, such as, CPU Cores per socket, CPU Frequency (MHz), etc. Click on one of these characteristics and the list will be reordered based on that criteria. The results don't affect the modeling, but will be visible when you complete the Document Generation later in the project. Below is a screen showing the Field Chooser popup menu.

Candidate Selection

Candidate Selection is an X_Factor module that uses a set of weighted values to generate an Xcedex Readiness Algorithm (XRA) score. This score is used by X-Factor to determine which workloads (hosts) are candidates for virtualization. This data is fed into the Host Modeling module which generates a model for your virtualized environment.

Topics covered here include:

• Screen Overview
• Xcedex Readiness Algorithm (XRA) Score
• Candidate Modeling Using Default Weighted Values
• Modifying the Weighted Values to Influence the XRA Score
• Modifying the Overall Value to influence the XRA Score
• Manually Including/Excluding Candidates.

Screen Overview

The Candidate Selection screen contains several "weighted values" that influence the XRA score. You may use default values for modeling or, depending on your business requirements, modify them to get the results that meet your needs. The weighted values are:

• Overall Value
• Memory Weight
• CPU Weight
• Disk Weight
• Network Weight

Starting from left to right, please note these values on the screen below. How to modify these values and what impact they have on Candidate Selection will be discussed in more detail later in this section.

Also on this screen you will see a "Model" button underneath Overall Value and two drop down lists called "Results" and "Candidate List". These will be discussed later in this document.

Xcedex Readiness Algorithm (XRA) Score

The XRA score is at the foundation of the quantitative candidate selection process. The proprietary algorithm assigns a score to each of the four core resources (CPU, memory, network, disk), adding them up to provide the XRA score. Servers with an XRA score higher than the set threshold of 100 are excluded as candidates. Servers with a score lower than 100 are included. On the example below you can see that the Factored Score of 266.14 would cause this candidate to be excluded as a virtualization candidate.

An Example XRA Score:

Any of the core 4 criteria may reach a threshold that may result in a candidate being excluded.

Common Candidate Elimination Thresholds

1. Memory utilization greater that 3 GB.
2. Computing power (MHz) greater than 2000 MHz
3. Network usage greater than 300 Mb/sec
4. Disk I/Ops greater than 2000

Scoring and Candidate Selection

The score is NOT meant to identify servers that cannot be virtualized. Rather, the score represents a relative “cost” of each workload within a company’s environment for virtualization. That is, even workloads with a high score can be virtualized, however the cost would be evident in either a higher number of physical machines required for the project or in increased resources (CPU, Memory, etc.) required per physical machine. This list of servers then, with their associated relative costs, can be used to prioritize workloads for virtualization and to design infrastructure accordingly.

Manual Adjustments to the XRA Recommended Candidates

The algorithm has set defaults for the four core resources and the overall value. These defaults create a Factored Score that is used to generate a virtualization plan with the least risk and lowest cost. However, these defaults may be manually changed based on user criteria to cause more or fewer workloads to be included in the virtualization plan. Also you may manually include or exclude candidates.

Automatic Exclusions

Servers with an XRA score above 100 will be automatically excluded as candidates.

Automatic Inclusions

Servers with an XRA score below the 100 will be automatically included as candidates.

Manual Exclusions

It is often necessary to excluded candidates from the project for reasons outside of its performance profile.  Perhaps policy dictates production domain controllers or SQL servers are physicals.  Manual exclusions are an easy way for you to implement and track your business rules.

Manual Inclusions

Candidates excluded for performance reasons may be included in the planning stage to ensure the Scenario is built with their workloads in mind. Virtualizing servers that fall outside the typical measure may occur for many reasons associated with your business. For example you could include in the model a SQL server running 1:1 as a VM to see how it affects the virtual infrastructure plan and make business policy based on this data.

Incomplete Data

This means that no data was captured on these hosts during the data collection process. The hosts were included in the inventory provided to CollectorX, but it was unable to interrogate them. Typically this means that these hosts were either offline, removed from service or an improper IP address was associated with them.

Candidate Modeling Using Default Weighted Values

The default values are based on an algorithm that was developed by Xcedex over hundreds of virtualization planning engagements. These values will enable X-Factor to produce a plan (candidates) that provides the least amount risk and the lowest cost. On the screen below you see Memory, CPU, Disk and Network Weight have a value of "50". This is their default and means that they are weighted against each other equally to create an individual resource score.

Overall Value contains a number of 100. This value is a percent and means that X-Factor will add together 100% of the individual resource (CPU, Memory, etc.) scores to generate the XRA score. As seen by the previous example, the XRA score is created by adding each of the resource scores together.

To model your candidate list using the defaults, click on the "Model" button above.

X-Factor will now calculate the score for each resource, create an XRA score and either Include or Exclude a candidate for virtualization. Under the "Results" bar below, you can see that X_Factor produces a Pie Chart that summarizes the disposition of the candidates. (10) were Automatically Excluded, (42) were Automatically Included, (15) had Incomplete Data, (0) were Manually Excluded, and (0) were Manually Included.

Below the Pie Chart is a bar called "Candidate List". Clicking on this Bar will show all the candidates, their scores and whether or not they were selected (Included) for virtualization planning. The FactoredScore for host, FIRE-APPS10 is highlighted in pink and indicates that the score is over 100 and that it has been automatically excluded as a candidate.

If you click on Details to the left of the FIRE-APPS10 a popup will appear that shows the individual resource performance data and the factored score for each.

Details for host FIRE-APPS10

Be sure and use the scroll bar on the popup to see all the resource data.

At the top of the Candidate list you have a series of column headings. You can sort the list by any column in either ascending or descending order. The black arrow within the column heading indicates the direction of the sort. The list below has been sorted by "FactoredScore" so that the Automatically Excluded hosts can be easily reviewed.

On the bottom of each page of the Candidate List you will see an "Export" button. By clicking this button you may export the entire list to an Excel file and save it to your PC. See below.

Modifying the Weighted Values to Influence the XRA Score.

In the first modeling example, we used the default values for each resource weight to generate the XRA scores and determine which hosts were included/excluded for modeling. In that example (10) hosts were Automatically Excluded.

You may change the weighted value for a given resource. This in turn will affect the XRA score for that host. In the example below I increased the "CPU Weight" value from 50 to 100. This tells X_Factor to calculate the CPU Score at a higher proportionate rate thus increasing its score. For every 10 point change to a weighted value (from the default of 50) will change that resource's score by 20%. In this case, raising the CPU weight to 100 (doubling the CPU score) caused (2) more hosts to be Automatically Excluded for a total (12) hosts. Inversely, you may lower the weighted values which may cause more candidates to be Automatically Included.

You may change the weight for each resource individually or two or more at the same time. Feel free to try different values to see how they impact the Candidate Selection.

Modifying the Overall Value to influence the XRA Score

The "Overall Value" is what percent of the XRA score should be used for candidate selection. The default is 100%. Changing this value will change the XRA score.

After running a candidate selection model using default values you may have several hosts that fall above the factored score threshold of 100 and excluded as candidates. Manipulating the Overall Value will allow you to include more candidates, and therefore, understand how higher utilized hosts will impact the virtualization plan. Below is candidate selection using the defaults. Note the value of each resource's score, the Factored Score and the number of included candidates.

On the screen below the Candidate Selection model has been run using an Overall Score of 50. Note the individual resource scores remain the same, but the FactoredScore has been reduced by 50%. from the value on screen above. Also note that (7) more hosts were included as candidates. Only 3 hosts remain above a FactoredScore of 100.

Manually Including/Excluding Candidates.

Aside from manipulating the weighted values to change which hosts are included/excluded as candidates, you can also manually their status. On the screen below I have sorted the candidate list by FactoredScore showing the score from highest to lowest. This view shows me the excluded candidates. To the right of the FactoredScore column is a column called "Manual". In this column you will see a check box next to each FactoredScore.

For demonstration purposes let's assume that all hosts starting with S037xxxx are to be included in the virtualization plan regardless of their score. To manually include them, click on the corresponding check box and you will see a dialogue box that asks whether you want to include or exclude the host. In this case we want to include each host starting with S037xxxx. On the screen below, each of these hosts has been manually included. The column to right of Manual is the "Reason" column. You may click on the corresponding box and enter the reason for manual inclusion. This is useful for documentation and tracking purposes.

To clear any manual changes, click on the check mark. This will clear the manual change and the corresponding reason.

Creating Custom Templates

1. Log on to X-Factor, highlight a project and click "Select Project". You will then be directed to the Planner screen.
   
2. On the Planner Screen click on "Data Visualization" on the left side of the screen to access a drop down menu. Then click on "Report Elements".
   
3. You now see a screen labeled Charts, Tables, & Text (above). By clicking on the various elements under Charts, Tables & Text you will see a layout of what X_Factor will generate. You will also see the specific "tag" associated with each. Determine the elements you would like to include in your own template.
4. For the example below, I used data from a sample project and clicked on "Candidate Selection Results" to see what the Chart would look like in my own documentation. Please note where the "tag" is shown.
   
5. After you determine the "tags" you wish to include in your MS Word document you simply enter the tag label exactly as it is shown (including the <> alligators) anywhere in your document.
6. Next you need to upload your document into the Project Documents section of X_Factor. On the Planner screen, click on "Data Visualization" and then click on "Documents". A Documents Template screen will appear.
   
   This screen will display all existing Xcedex templates as well as any custom templates that have been uploaded. From this screen you can upload your custom template(s).
7. Then click on "Add Template" and a pop-up screen will appear as shown below. Simply click "Browse", select your custom template and then click "Submit File". Your template will be uploaded into the Documents section of your project.
   
8. At this point your template is ready to be populated with the results of your data collection and modeling. To do this, navigate to your Document Template screen. Click the desired template and then click on "Generate."
   
9. To locate your generated document, click on the "Previously Generated Documents" bar. Your document will be displayed below the bar.
   
10. The final step here is to copy the generated document to your own PC. Do this by clicking on "Download" next to the desired generated document. A file download dialogue box will appear. Click on "Save" and specify the location where you want to save the file.
    

Combining Standard Template Elements with a Custom Template

This method involves moving generated elements from an X_Factor Template to a user-customized document.

1. Log on to X-Factor, highlight a project and click "Select Project". You will then be directed to the Planner screen.
   
2. On the project planner screen click on "Data Visualization" to access a drop down menu. Then click on "Documents". A "Documents Template" screen will appear.
   
3. On this page click on "Generate All Items.doc" and then click on 'Generate"
   
4. A dialogue box called "Create New Document" will appear. In this box you fill in the desired document (file) name and an optional description then click on "OK".
   
5. You will then see a dialogue box prompted to "Open" or "Save" the document. If you select "Open" the document will open as a Word document and allows you to review the generated elements immediately. Or you can click on "Save" to save it on your PC and review it later.
   

This Word document will contain pie charts, tables, dashboards and text summaries of all the modeled elements. You can review the elements and choose which ones you would like to in-clude in your custom document simply by doing a "copy and paste" operation.

Note: If a desired element is missing, please refer to "Creating Custom Templates" to deter-mine how to insert individual elements into your Word document.

Using Standard X_Factor Templates

1. Log on to X-Factor, highlight a project and click "Select Project". You will then be directed to the Planner screen.
   
2. On the project planner screen click on "Data Visualization" to access a drop down menu. Then click on "Documents". A "Documents Template" screen will appear.
   
   On this screen you will see a list of X_Factor/Xcedex Templates that generate Word documents that show information ranging from graphical item summaries to detailed documents that con-tain text and graphics. We recommend that you generate and review these templates to de-termine which better suits your needs. If none of these templates fully meet your needs, you can "Copy and Paste" the pertinent information or graphics into your own document. The steps below will show you how to generate and review these templates.
3. Select a template and click on "Generate".
   
4. In the "Create New Document" dialogue box enter a Document Name and an optional Description and click on "OK"
   
5. At this point you can either save the document to your PC or open it to review its relevance to your needs. Remember, you can "Copy and Paste" specific text or graphic elements in a template document to you own custom Word document.