Creating Insights
Insights are displayed in the Insights widget on the Summary page of your default view. They identify opportunities where you can take action to optimize your IT infrastructure. Insights alert you to situations that are either costing you money or are potentially harming your environment in some other way.
Insights are divided into four categories:
- Cost Savings opportunities
- Unusual changes in usage or cost
- Optimization opportunities
- Personalized Insights
In each of the first three categories, HPE Consumption Analytics portal provides several ready-to-use Insights based on predefined rules. You can also create your own using the delivered rule templates (see Creating an Insight). Any Insights defined by your administrator specifically for your organization will display under the Personalized category.
HPE GreenLake Accounts
The illustration that follows shows the Insights widget on the Summary page in the global Public Cloud View, for a public cloud account. The Summary page in the HPE GreenLake Customer View is shown in Working with the HPE GreenLake Customer View.
The summary of your active Insights in your default view shows the number of Insight alerts in each category. The bar at the top of the widget shows the amount you can potentially save by resolving these alerts, if over $5000 per month; otherwise it shows the total number of Insight alerts you have.
There are different default Insights for HPE GreenLake products and public cloud accounts. For a public cloud account, you may see an alert when the cost for a department increases suddenly, or when you're paying for a VM that no one is using. For an HPE GreenLake product, Insights are targeted more toward capacity than cost; you may see an alert when a resource is nearing capacity, or when there's been an unusual increase in installed capacity for a given device.
Understanding your Insights
All Insights have an absolute threshold, against which a raw change in the data is compared. If the absolute threshold is not crossed, no alert is generated. Most Insights also have a relative threshold, against which a percentage change is compared. In this case, both the relative and absolute thresholds must be crossed to raise an alert.
Relative thresholds are useful to limit alerts to those that are truly significant. For example, in a rule for monitoring changes in the cost of service categories, the relative threshold is 10% and the absolute threshold is 10. If the cost of the Messaging service increases by a relative amount of 46%, and a raw amount of $0.20, no alert is sent. While the 46% increase seems large, it only represents an additional $0.20 in actual cost, which is far below the absolute threshold.
When you drill into a row in your Insights widget, the Insights Drill Down overlay shows a list of your Insights in the left navigation pane. In the main pane, the detail for the selected Insight displays. This includes a visualization of your actual consumption data, which contains one or more charts:
If you don't want to see the visualization, click HIDE VISUALIZATION. HPE Consumption Analytics portal will remember your setting and hide visualizations for all Insights you view. To view the visualization again, click SHOW VISUALIZATION.
The second part of the Insights Drill Down is a table that shows rows of your actual consumption data. The Insight Status column lists the alert status of each row of data:
- Alert: The Alert status is highlighted in red. Data has crossed all thresholds and has generated an alert.
- Approaching: The Approaching status is highlighted in yellow. Data is within 10% of the thresholds.
- Normal: The Normal status is highlighted in green. Data is well below the thresholds.
For example, the following table shows data for a rule that has a relative threshold of a 5% increase in Daily Cost, and an absolute threshold of $1.
The first row has crossed both thresholds, as seen in the Change Percent and Change Amount columns, and is therefore in the Alert status. Other rows have crossed the absolute threshold by a wide margin, but remain in the Normal status because the relative threshold has not been crossed.
The table in the Insights Drill Down includes metrics to help you evaluate the situation and determine what action to take. For example:
- Average in % - The sum of the daily values for utilization percentage, divided by the number of days monitored.
- Average Network In (MB) - The average volume of incoming network traffic, expressed in MB.
- Average Network Out (MB) - The average volume of outgoing network traffic, expressed in MB.
- Days Crossed - The number of days within the configured time frame that the thresholds have been crossed.
- Days Approaching - The number of days within the configured time frame that consumption is within 10% of the threshold(s).
For public cloud account Insights grouped by Resource ID, you can click the launch icon in the Resources field to open those resources directly in the AWS or Azure portal (in a new browser tab).
Understanding rule templates
As described earlier, Insights are rule-based. Each rule template contains a rule definition that specifies what is monitored and what generates alerts. Rule definitions are constructed as sentences, with blanks you complete to define your Insights. The structure of the sentence and the entries required to define it depend on the type of Insight.
The default rule templates are fully defined with values, making it easier for you to quickly modify a rule when you create a new Insight. For example, the following image shows the rule definition included in the Numeric change for a group rule template:
This default rule raises an alert when the Cost for any Service Category on the most recent 1 day is higher than the previous daily average by 10% or more, and the absolute change in cost is at least 10 EUR. Data from the most recent 2 weeks is used for the daily average. When you create a new Insight and select this rule, you can change any of the underlined variables by entering a different value in the field, or choosing one from its selection list.
If a threshold is set to a value of 0, it is not used to determine whether to generate an alert.
You can use the delivered rule templates to create new Insights quickly and apply them to your consumption data. Because many of the templates are similar, it's helpful to understand them in more depth before using them to create new Insights. There are three standard templates on which related default templates are based:
- Numeric change for a group
- Numeric level sustained for a group
- Outlier
These are described under the next three headings.
Numeric change for a group
The Numeric change for a group template is a standard rule template that monitors the percentage and raw change (either increasing or decreasing) of a metric across a category. This default rule raises an alert when the daily average of the Cost metric for any Service Category is higher than the previous daily average by 10% and the absolute change is at least 10. Data from the most recent 2 weeks is used for the daily average.
This template provides great versatility for various metrics, and is used as the basis for the following default rule templates:
| Rule Template | Description |
|---|---|
| Cost increase for a group | This default rule raises an alert when the daily average of the Cost metric for any Service Category is higher than the previous daily average by 10% and the absolute change is at least 1. Data from the most recent 2 weeks is used for the daily average. |
| CPU utilization change for a group | This default rule raises an alert when the daily average of the CPU Utilization metric for any Service Category is higher than the previous daily average by 10%, regardless of the absolute change (this template does not have an absolute threshold). Data from the most recent 2 weeks is used for the daily average. |
| Quantity change for a group | This default rule raises an alert when the daily average of the Quantity metric for any Service Category is higher than the previous daily average by 10%, regardless of the absolute change (this template does not have an absolute threshold). Data from the most recent 2 weeks is used for the daily average. |
Numeric level sustained for a group
The Numeric level sustained for a group template is a standard rule template that, for the group you specify, monitors whether a numeric field value remained above or below a threshold for a length of time. This default rule raises an alert when the daily average of the Cost metric for any Resource ID has been above 0 each day for the past week. Based on this template, HPE Consumption Analytics portal includes templates for sustained utilization and underutilized resources.
HPE Consumption Analytics portal includes the following templates for sustained utilization. While the default settings are for metric values below specified levels (underutilization), you can change these templates to monitor for values above specified levels (overutilization).
| Rule Template | Description |
|---|---|
| VM with sustained utilization | This template monitors VM CPU utilization to help you identify underutilized or overutilized VMs. The default rule is set for underutilized VMs; it raises an alert when the daily average of the CPU Utilization metric for any VM has been below 2%, while the daily total network traffic remains below 10 MB each day for the past 5 days. |
| EBS volume with sustained utilization | This template applies only to AWS resources. It monitors EBS volumes for read and write traffic that remains above or below thresholds for a length of time. The default rule raises an alert when the daily averages of both read activity and write activity for any EBS volume have been at or below 0 MB each day for the past 4 days. |
| ElastiCache cluster with sustained utilization | This template applies only to AWS resources. It monitors ElastiCache hosts for read and write traffic that remain above or below thresholds for a length of time. The default rule raises an alert when the daily averages of both read activity and write activity for any ElastiCache host have been at or below 0 MB each day for the past 4 days. |
| RDS instance with sustained utilization | This template applies only to AWS resources. It monitors RDS instances for read and write traffic that remain above or below thresholds for a length of time. The default rule raises an alert when the daily averages of both read activity and write activity for any RDS instance have been at or below 0 MB each day for the past 4 days. |
| Redshift node with sustained utilization | This template applies only to AWS resources. It monitors Redshift nodes for traffic that remains above or below thresholds for a length of time. This default rule raises an alert when the daily averages of both inbound and outbound traffic for any Redshift node have been at or below 0 each day for the past 4 days. |
HPE Consumption Analytics portal includes the following delivered templates to help you identify underutilized resources, which are candidates for deletion. These templates are very similar to the templates for sustained utilization. The key difference is that these templates are always configured to monitor for underutilization. You cannot change the At or below setting in these templates.
| Rule Template | Description |
|---|---|
| Underutilized VM | This template monitors VM CPU utilization to help you identify underutilized VMs. This default rule raises an alert when the daily average of the CPU Utilization metric for any VM has been below 1%, while the daily total network traffic remains below 3 MB for each day for the past 5 days. |
| Underutilized object storage bucket | This template applies only to AWS and Azure resources. It monitors AWS S3 (Standard class) and Azure Blob (Hot tier) for read and write operations that remain at or below thresholds for a length of time. The number of read and write operations correspond to Request-Tier1 and Request-Tier2 usage types for AWS S3 and Read Operations and Write Operations for Azure Blob. |
| Underutilized EBS volume | This template applies only to AWS resources. It monitors EBS volumes for read and write traffic that remain below thresholds for a length of time. This default rule raises an alert when the daily averages of both read activity and write activity for any EBS volume have been at or below 1 MB each day for the past 7 days. |
| Underutilized ElastiCache cluster | This template applies only to AWS resources. It monitors ElastiCache clusters for read and write traffic that remain below thresholds for a length of time. This default rule raises an alert when the daily averages of both read activity and write activity for any ElastiCache cluster have been at or below 0 MB each day for the past 4 days. |
| Underutilized RDS instance | This template applies only to AWS resources. It monitors RDS instances for read and write traffic that remain below thresholds for a length of time. This default rule raises an alert when the daily averages of both read activity and write activity for any RDS instance have been at or below 1 MB each day for the past 7 days. |
| Underutilized Redshift cluster | This template applies only to AWS resources. It monitors Redshift clusters for traffic that remains below thresholds for a length of time. This default rule raises an alert when the daily averages of both inbound and outbound traffic for any Redshift cluster have been at or below 0 MB/second each day for the past 4 days. |
Outlier
The Outlier template is a standard rule template that monitors whether one group is different from its peers. This is useful for finding the "bad actors" in your organization who spend or use more than others. The default rule raises an alert when the daily average of the Cost metric for any Account Owner for the past 5 days is 50% higher than the average of all other Account Owners and the absolute difference is at least 100.
Each rule also lets you filter the data you want to monitor. For example, you might want an Insight to apply only to data from a specific provider or region.
Results Drilldown
The Results Drilldown option under Insight Configuration provides an optional drill-down functionality for your Insights. This allows you to select a report to view within the Insight. This is useful to help you determine if an action needs to be taken based on the recommendations. From the selection list, select a relevant report from the MY REPORTS group or a shared report from the SHARED WITH ME group.
After you've selected a report, in the Complete Result Set, click the name of the report in the Drilldown column to open your results
The results of the drilldown will display in an overlay.
When you export data in PDF, Excel, and CSV, the report maintains the drilldown layout.
Video overview
The following video shows you how to configure the default Insights included with the HPE Consumption Analytics platform.
Creating an Insight
The following steps show you how to create custom insights in the Insights page, using the rule templates.
To create a custom Insight
- Click Insights in the left navigation.
- On the Insights page, click the + icon .
The New Insight overlay displays. - In the Rule Template field, select a rule template that provides a good starting point for your rule. For more information about the default rules, see Rule templates.
- In the Rule Definition area, change the variables to field names and values as appropriate for the rule you want to create. As you define the Insight rule, its visualization and table will update.
You can adjust variables in the following ways:
- Variables that have a fixed selection of choices, such as date ranges, are presented as drop-down fields with an arrow next to the variable. Click the field and select a value.
- For variables that require a field name, click the Close icon next to the variable to remove the existing field name and open a selection list for a replacement field name. Alternatively, you can click the value, backspace to remove the existing text, and begin typing a new field name. As you edit the field, HPE Consumption Analytics portal opens a selection list with values that match your text.
- For variables that require a numeric value, click the value and begin typing.
- If you want your Insight to apply only to a subset of available data, click Add Filter. For more information about using filters, see Filtering data.
- Save the Insight:
- Click Save.
The Save Insight dialog box appears.
- Enter a Name, and (optionally) edit the Description provided by HPE Consumption Analytics portal.
- In the Recommendation field, enter your recommended action when this Insight generates an alert.
- Select a Category, which define where the Insight will appear in the Insights summary widget in your default view.
- In the Share With area, add one or more groups and users that can use this Insight. If you want all users to use it, select Share With All Users.
- Click Save. Your new Insight is added to the list of Insights.
- Click Save.
Video example
The following video shows you how to create new Insights using a template:
What to do after creating Insights
Create user accounts and groups for the people that need to view usage and cost data. For more information, see Permissions.