8.8 Designing an Aggregated View

Viewpoint enables you create new aggregated views and edit the provided aggregated views or aggregated views you have created. This topic identifies how aggregated view components are organized and the resources available when designing an aggregated view.

See 8.5 Creating an Aggregated View for instructions on how to create an aggregated view.

See 8.7 Editing an Aggregated View for instructions on how to edit an aggregated view.

In this topic:

• 8.8.1 Aggregated View Information
• 8.8.2 Aggregated View Permissions
• 8.8.3 Using the Pipeline Designer

8.8.1 Aggregated View Information

There are two fields at the top of the Create and Edit Aggregated View pages where you can enter identifying information about the aggregated view you are designing:

Name – The name of the aggregated view. This is a plain text field, which may contain uppercase or lowercase letters, numbers, spaces, or special characters.

Description – A description of the aggregated view. May also be used to provide instructions to report designers on how to use the aggregated view to create reports.

8.8.2 Aggregated View Permissions

You can restrict who has access to your aggregated view. This is done using the Permissions area at the top right of the page.

The following image is an example of the Permissions area.

Click to enlarge

Using this area you can restrict access by:

• Users. In the Users field, type the IDs of the users. Viewpoint will check if the users you added are valid; that is, has a valid operating system account. If the user is valid, the user name turns green; otherwise it turns red. Use "ALL" to remove restrictions.
• Group associations. In the Groups field, type the IDs of the groups. Viewpoint will check if the groups you added are valid; that is, has a valid operating system account. If the group is valid, the group ID turns green; otherwise it turns red. Use "ALL" to remove restrictions.
• Account associations. In the Accounts field, select from the available accounts listed in the drop-down. Leave blank to remove restrictions.

8.8.3 Using the Pipeline Designer

When you create a new aggregated view, the Pipeline Designer for the view appears as shown below:

Click to enlarge

The available node types are represented by icons on the left side of the page (source, filter, etc.). The current pipeline design is represented by icons placed in the design area on the right side of the pipeline designer. To increase the size of the design area, click and drag the icon on the lower right corner of the design area.

An aggregated view consists of a connected series of nodes (the "pipeline") that operate on data from a data source to produce a database layout. The Pipeline Designer places nodes representing the data source and database layout for you in a new pipeline. To complete the design of your aggregated view pipeline, place and connect the nodes that operate of the data from the data source to produce your desired database layout.

There are a few basic steps in placing a node in a pipeline:

1. Drag a node from the left side of the page and drop it in the design area.
2. Connect the node to the previous node in the pipeline.
3. Edit the node's properties.
4. Add a comment describing the node (optional).
5. Click Apply to apply the properties to the node.

A finished aggregated view pipeline is shown below:

Click to enlarge

8.8.3.A Pipeline Data

Data flowing through the aggregated view pipeline is represented using JavaScript Object Notation (JSON). JSON is a standard format for data interchange, designed to be easily readable.

The data flowing through a pipeline consists of a stream of timestamped messages:

{
   "msg": {
      //message data
   },
   "ts": <timestamp of the message>
}

The structure and content of the data portion of each message depends on whether the data source is a job, reservation, cluster sample, or job state journal stream.

You can view the structure of the data stream at any point in a pipeline design by clicking on a node, then clicking the Inpsect Incoming JSON structure link below the design area. For example, in the pipeline design shown above, if you click the Filter node, then click Inspect Incoming JSON structure, you will see the structure of the stream of job messages passed from the job data source to the Filter node.

Click to enlarge

8.8.3.B Pipeline Nodes

The nodes in a pipeline operate on and transform the incoming data stream. A description of the operation of each node type is given in the table below.

{
   "msg": {
      "key": <key by which grouping is performed>
      "value": <result of the aggregate function>
   },
   "ts": <timestamp of the message>
}

//JSON before flattening where key="msg.arr", type=integer, value=1
{
   "msg": {
      "arr": ["one", "two"]
   },
   "ts": 123
}

//JSON after flattening
{
   "msg": {
      "key": "one",
      "value": 1
   },
   "ts": 123
},
{
   "msg": {
      "key": "two",
      "value": 1
   },
   "ts": 123
}

8.8.3.C Saving an Aggregated View Design

When you have finished designing your aggregated view, click SAVE AND CLOSE to save the aggregated view and close the page or click SAVE AND RUN to save and view the aggregated view.

Related Topics 

• 1.1 Designing an Aggregated View
• 1.1 Designing an Aggregated View
• 8.4 Viewing an Aggregated View Pipeline
• 8.5 Creating an Aggregated View
• 8.6 Importing an Aggregated View
• 8.7 Editing an Aggregated View
• 8.9 Provided Aggregated Views
• Chapter 8 Reporting