- Rollup rules: Lists your rollup rules, with shaping impact data for each rule. This tab is selected by default.
- Mapping rules: Lists your mapping rules. This tab appears only when your tenant has at least one mapping rule. You can’t create mapping rules from this page.
Search and filter rules
Use the Search rules field to find a particular rule or group of rules. The search matches each rule’s name, slug, filters, status, policy and labels, output metric name, and last-modified date. Dates match the format shown in the Last modified column. Searches aren’t case-sensitive. Separate multiple terms with spaces to return only rules that match every term. For example,kubelet discard returns rules that contain
both kubelet and discard.
On the Rollup rules tab, use the Status menu to show All rules, or only
rules in Preview or Enabled mode.
Rollup rules table
The Rollup rules table includes the following columns. To display a description of a column, hold the pointer over the information icon in the column header.- Status: The mode the rule is in, either Preview or Enabled. To learn about preview mode, see Preview aggregation rule impact.
- Name: The rule name.
- Filter: The names of the metrics or labels the rule matches. Filters can include both Prometheus and Graphite metrics.
- Policy: The rule’s label policy. Keep labels indicates the rule preserves the listed labels and discards all others. Discard labels indicates the rule removes the listed labels and keeps all others. Drop raw indicates the rule deletes the raw metrics it matches.
- Metric name: The name of the output metric that the rule creates and persists
to the database. In the output name,
{{ .MetricName }}represents the name of the input metric. - Matched writes, Persisted impact, Efficiency, and Utility score: The columns that make up the rule’s shaping impact.
- Last modified: The last time the rule was updated.
Shaping impact
Shaping impact data shows the individual effect of each existing rule on your metrics traffic, which helps you manage your traffic shape. The following columns describe the shaping impact of a rule:- Matched writes: The raw number of incoming data points per second matching the rule. If your organization uses consumption-based pricing, this column is named Matched datapoints.
- Persisted impact: The change in total persisted data points per second as a result of the rule. This column doesn’t appear if your organization uses consumption-based pricing.
- Efficiency: A measure of how effective the rule is in aggregating incoming data points, expressed as a percentage reduction in data points per second from incoming to aggregated data.
- Utility score: An aggregate number that indicates the relative usefulness of the output metric, determined by the number of references, executions, and unique users. A higher score means users include the output metric in their workflows. To learn more, see Telemetry Usage Analyzer.
- To reduce persisted data, look for rules that have low utility but add persisted data. Use the Least Valuable (by Persisted Impact) sort to find them. This sort isn’t available if your organization uses consumption-based pricing.
- To clean up ineffective rules, look for rules that aren’t aggregating data, or don’t match incoming data points effectively. Use the Least Effective sort to find them.
Sort your data
On the Rollup rules tab, use the sorting menu to order rules by their shaping impact. The menu is set to Custom by default, and includes these options:- Least Effective: Lists rules with the lowest rule efficiency first.
- Most Effective: Lists rules with the highest rule efficiency first.
- Custom: Sorts by the column you select in the table header.
- Least Valuable (by Matched Impact): Lists rules with the lowest output metric utility relative to their matched writes first.
- Most Valuable (by Matched Impact): Lists rules with the highest output metric utility relative to their matched writes first.
- Least Valuable (by Persisted Impact): Lists rules that add persisted data and have low output metric utility first.
- Most Valuable (by Persisted Impact): Lists rules that conserve persisted data and have high output metric utility first.
Mapping rules table
The Mapping rules tab includes a table with the following columns:- Name: The rule name.
- Filter: The names of the metrics or labels the rule matches.
- Policy: The resolution at which the rule stores matched metrics, shown as
sample <resolution>. Drop raw indicates the rule deletes the raw metrics it matches. - Last modified: The last time the rule was updated.
Add aggregation rules
On the Rollup rules tab, click Create rule to open the Create Aggregation Rule drawer. Use the rollup rule creation procedure to define the rule. Cortex XCOR doesn’t save rules you define in the drawer. Define the rule on the Visual editor tab, and then select the Code config tab to copy the configuration and apply it with Terraform or Chronoctl.Rule details
Click any rule in a table to open its details drawer. Click Previous or Next to move between rules of the same type without closing the drawer. The drawer header contains the following controls:- Add comment or View comments: Opens the comments for the rule. When the rule has comments, the button label includes the number of comments.
- View incoming metrics: Opens the Live Telemetry Analyzer to analyze the rule’s incoming data.
- Edit: For rollup rules, opens the rule in the editor.
- View Configuration: For mapping rules, displays the rule’s code configuration.
Rule configuration
The Rule configuration section lists the rule’s settings, including its status, last update time, filters, and whether it drops raw data. For rollup rules, the section also lists the slug, label policy, output metric name, metric type, aggregation, and interval. To view the code configuration for a rule:- For rollup rules, click Edit, and then select the Code config tab.
- For mapping rules, click View Configuration.
Shaping impact graphs
For rollup rules, the Shaping impact section displays a current value and a historical graph for each of the following measures:- Matched writes: The raw number of incoming data points per second (DPPS) matching the rule. If your organization uses consumption-based pricing, this measure is named Matched datapoints.
- Persisted impact: The change in total persisted data points per second as a result of the rule, as a percentage. The current value also displays the number of DPPS added or DPPS conserved. This measure doesn’t appear if your organization uses consumption-based pricing.
- Rule efficiency: The percentage reduction in data points per second from incoming to aggregated data.
- Utility score: The output metric utility score over time.