|CMDB Analyzer User Guide|
This user guide describes CMDB Analyzer version 2.0 and later.
Welcome to use CMDB Analyzer! After you download and install CMDB Analyzer, you can start to analyze intuitively your complex CMDB CIs and their dependencies and relationships with much greater clarity.
Installing CMDB AnalyzerIf you download the zip file version then there is no install process. Just unzip the file and start to run. Otherwise follow the instruction of the installer.
System RequirementsCMDB Analyzer is written in 100% Java. The memory requirement depends on the size of CMDB data you are viewing. Without loading full instance attributes, CMDB Analyzer uses about 100M memory to load 10k CIs with 10K relationships.
change the application maximum heap size, edit file cmdbAnz.l4j.ini.
CMDB Analyzer uses CMDB SDK version 2.0.
The following diagram describes function areas in CMDB Analyzer main window:
A workspace is a directory that contains CMDB Analyzer projects. Only one workspace is activated but you can switch workspaces.
To create a CMDB project, you have to specify a data source. The data source can be either exported xml data files or a live CMDB server.
Create a project to load CMDB xml data file
You need to prepare three types of files to create a file project.
If your instance file already contains UI instances, you don't need a separate UI file.
CMDB xml data files can be generated using CMDB SDK command line utility "cmdbdriver.exe", which can be found in the CMDB SDK bin directory. You need to export both class definitions and instances in order to define a file data source. The following steps demonstrate how to export CMDB data using cmdbdriver.exe:
There are several other ways to generate instance file:
Once you have the CMDB data files ready, enter them in the project definition window:
If you enter a directory name instead of a particular file name, CMDB Analyzer will load all relevant files found in that directory. If your instances file already contains UIComponent instances, you can just leave "UIComponent File Or Directory" empty.
Create project to connect to a live CMDB server
Enter ARS login information in the project definition window:
If you uncheck "Save Password" option, the password will not be saved - you have to enter the password every time you open the project.
Load All Instance Attributes
Normally you should not select this option because loading all instance attributes takes long time and consumes large memory. You can always retrieve all attributes of an instance by clicking "refresh" button in the instance detail panel.
Tip: Uncheck "Load All Instance Attributes" to reduce the memory consumption. As a result, only a few essential attributes, such as Name and InstanceId, will be loaded and attached to CI instances and relationships, but the memory saving is substantial. For example, memory usage goes down from 150M to 60M by unloading attributes. You have to refresh the data source after making the change.
Retrieve Exact Icon
If you haven't modified BMC.CORE.CONFIG:BMC_UIComponent Form which defines class icon information, or you know you have only one icon defined for each CI class, then don't check this option. On the other hand, if you have qualifications defined in UIComponent instances, and you want to correctly display class icons (for example, you have both desktop icon and laptop icon defined for BMC_ComputerSystem), you can check this option, but beware that you may experience downgraded performance as a result of selection this option.
Tip: To change project settings after project has been created, right-mouse click on a project in Project Panel and then select "Preference...".
Project Panel organizes projects and their views in a workspace. The followings are the basic operations to manage projects and views.
System configuration views are only available for projects with live CMDB connection.
Following is a screenshot of Reconciliation View:
CMDB Analyzer displays all datasets (including overlay dataset) available in a data source. In Project Panel, overlay dataset nodes are placed inside its master dataset node. Each dataset has its own CI views.
The following is the fully expanded toolbar seen in CDM View:
You can also use mouse wheel to zoom in and zoom out.
Instance Summary View reports the number of instances of each class.
CI View displays all CI instances and relationships of a particular dataset. CMDB Analyzer by default creates a CI View called "Main View" for each dataset. You can create duplicated views and views that only contain a subset of the data.
The following is the fully expanded toolbar seen in CI View:
Advanced mouse operations:
Graph Filter allows you to select desirable CIs and relationships to reduce the complexity of the graph. You can use the Graph Filter window to change the filter settings. Graph Filter applies only to a particular view - it won't reduce the instances number but just hides them. In the later section we will discuss the use of Instance Query - instance filters that actually reduces the instances loaded to a dataset.
Smart zoom allows you to spread out otherwise crowded CI instance nodes and relationship lines in the graph. You can use the mouse wheel or the zoom buttons in the toolbar to perform smart zoom.
You can use Object Magnifier to view parts of the graph in great detail, or use it as an overview window to have a bird view of the whole graph. The CIs and relationships in the Object Magnifier window are all alive - you can perform any mouse or keyboard actions as if you are working in the main view window.
Zoom out and turn magnifier as an overview:
Basic Object Magnifier Mouse Operations :
Relationship View gives you a clear local view of upstream and downstream relationships of a particular CI. You can drill up and drill down the graph by click the drill button or double click on a CI instance node in Relationship View.
Tabular View lists all visible instances in a table. The columns are configurable and you can export the table content to a CSV file which can be opened by Microsoft Excel.
Click "Choose Fields.." and you can choose table columns.
You can use Graphics Properties window to change graphics attributes of CI nodes, relationship lines, graph background and grid fill color. Once the CI View is saved these graphics properties are saved too.
You have three graph layouts to choose:
Top-to-bottom and left-to-right trees are good choices for laying out simple graph. Radial tree is more suitable when a large number of nodes are involved with complex connections between nodes. You might need to try all three layouts to select a best layout option.
There are two ways you can create a new CI View - Duplication and Subgraph.
Duplicating a CI View basically generates an exact copy of the view. To create a duplication, you can
You can also create views that show only a subset of CI instances and relationships in the dataset. We call this kind of view a subgraph.
Starting from a CI instance, you can create an outbound subgraph, an inbound subgraph, or an all-direction subgraph.
The screenshot below shows the difference of these three subgraph types.
To create a subgraph, right-mouse click on a CI instance node and select menu item "New Subgraph View...".
Create Tight Subgraph
A subgraph can be "loose" or "tight". "Loose" means there could be opposite branches in the subgraph. For example, an outbound subgraph (going downstream) has some upstream branches because some nodes have multiple parent nodes. "Tight" subgraph, on the other hand, can not have opposite branches. For example, a tight inbound subgraph (going upstream) should not have more than one outbound (downstream) path. The picture below shows the difference between a "loose" subgraph and a "tight" subgraph.
Defining Instance Search Filter for details about how to use Instance Filter Set Editor.
To edit dataset Instance Filter, right-mouse click on a dataset node. See screenshot below.
Instance Attribute Pane shows detailed information about the CI or relationship instance you select. To show or hide Instance Attribute Pane, select menu item "Show Attribute Pane" or "Hide Attribute Pane" in the CI popup menu, in the relationship popup menu or in the view popup menu.
Operations in Instance Attribute Pane:
Click "Snapshot" button on the CI View toolbar to take a snapshot of the view and copy it to clipboard. The snapshot image only covers the visible view window area.
Tip: Use "Export Image" option to export the whole graph to an image file.
To export the whole graph to a gif image file, click "Export" on the CI View toolbar and select "Export Image...".
To export the graph to a XML file, click "Export" on the CI View toolbar and select "Export Data...". Only visible CI instances and relationships are exported.
Tip: You can use the exported instance files as data source for a new project.
Open the Find Instance Window to search for a CI or a group of CIs by its name or instance ID. If you type in a partial string you will get a list of CIs whose names or IDs contain the string.
Another way to search for CIs is to use CI Tabular View. Click "CI Tabular View" on the CI View toolbar and you get a table of all visible CIs. Then you can sort the table and find the CIs you are looking for.
Click "Print" on the CI View toolbar to pop up the print preview window.
Options in the print preview window:
Change Diff Criteria
The default criteria to compare instances is the InstanceID. Click "Criteria" on the Diff View toolbar to change criteria setting. You can choose InstanceID, ReconciliationID or other instance attribute.
Tip: You can drag an instance (a row) in the Diff View result table and drop it to Instance Explorer.
Instance Explorer supports instance (both CI and relationship) drag&drop operation. You can drag instances from various sources, such as from CI View, Dataset Diff View and Instance Filter View. You can also directly enter instance ID by selecting menu item "Insert CI Instance" or "Insert Relationship Instance" in the view popup menu (right-mouse click on empty space).
Instance Explorer can combine instances from different datasets which have to be in the host project. You will get an error message if you try to add instances from other projects. Not like CI View, Instance Explore has no persistence - you can't save Instance Explorer. "Default Explorer" is the default Instance Explorer when the project is created. You can create additional Instance Explorer views by selecting "New Instance Explorer..." menu option on the Instance Explorer tree node (right-mouse click on the tree node).
CI Menu and Relationship menu
Instance Query Set is available only for server project.
A set of instance search criteria can be defined and saved as an Instance Query Set. The result instances are the combination of all the instances returned from the sequential execution of queries in the set. For example, you can define an Instance Query Set that first retrieves all instances from BMC_ComputerSystem, and then all instances from BMC_Processor, and finally retrieves all BMC_Component relationship instances. The result instances are presented in the CI instance table and the relationship instance table. You can drag a table row to Instance Explorer to explore the relationships. The screenshot below is the Query Set View.
Let's walk through the steps to create a query set.
There are four types of Instance Query you can create.
Create Class Query
A class query retrieves all instances of defined classes. To create a class query:
In the above screenshot, a class query is defined to retrieve all instances from dataset "BMC.ASSET".
Create Qualified Query
A qualified query retrieves instances from single CMDB Class with qualified class fields. To create a qualified query:
Create Graph QueryA graph query retrieves a branch of graph including both CI and Relationship instances. It starts from a single top CI instance to certain levels either upstream or downstream. The screenshot below defines a graph query that starts from a CI whose instanceId is "00001" and includes all levels of downstream relationships whose source CI class is BMC_BaseElement and destination CI class is also BMC_BaseElement.
Create SQL Query
A SQL query allows you to run database SQL statements to find CI instances and relationships. It is the most flexible and powerful Instance Query. Since CMDB Analyzer doesn't check SQL syntax for you, it's a good practice that you verify your query first and then copy the query to the query edit window. Because the result of the query must be either a list of CI instances or Relationships, the first part of the query "Select ... From" is done for you already. All you have to do is
The following example shows a SQL query that retrieves relationships whose CIs have "Company" field set to "bluelineG".
The "Table Name" button helps you to find the actual database table names via more readable form names. For example, it tells you that table "T65" is "BMC_BaseElement", and table "T122" is "BMC_BaseRelationship".
CI Aging Analysis gives detailed statistical information about when CIs in the production dataset are created, last modified and scanned.
CI Aging Report
CI Detailed Aging Report (Daily)