Cardinality management

As an administrator at New Relic, one of your tasks is to make sure you don't go beyond your limit of GB Ingested. One tool for tracking your ingest is the Cardinality Management UI. You'll find helpful graphs and tables explaining your ingest so you can take action if you're approaching limits.

Where can you manage cardinality?

In the New Relic UI, you can find this with other administration tools. Click your name in the lower left and go to Administration > Cardinality Management.

Account Cardinality Breakdown

If you're in an organization with multiple accounts, you can see a breakdown of cardinality by each account. As with most capabilities within New Relic, you can change the account you are currently viewing by selecting the account picker in the top right corner. The UI will display information about the selected account, as long as you have access to that account.

Usage over time

On the opening page of Cardinality Management, you can view the Usage over time chart:

This shows following:

• The limit that is currently at 5 million account cardinality.
• The current cardinality usage for the account, which is in the 2 to 4 million range. If there's no change in the value, then a new data point isn't reported. This can appear as an empty data point, but it's not data loss. This behavior may change as the feature is finalized.

Metrics table

In the Metrics table, you can see the breakdown of account cardinality by metric name:

This table shows the top 20 highest cardinality metrics in an account. The table provides the following details:

• The name of the metric
• The current overall usage of the metric (the ratio of current usage to current limit)
• The actual values of the current cardinality
• The current limit of that metric name

Metric cardinality breakdown

You can view a deeper layer of granularity in your metrics by clicking on any metric name. You can view a breakdown of the specific metric's cardinality details.

Metric and account cardinality bar charts

In this metric-specific view, you can see the current cardinality usage as two bar charts, one for Metric cardinality budget and one for Account cardinality budget (the overall account usage).

Usage over time (for metrics)

Below the bar charts is a chart of a timeseries query showing the metric cardinality usage versus the limit of that metric. Again, you can change the time-picker to view different time periods of this individual metric's cardinality data.

Attributes table

Under the timeseries, a table shows the attributes that are powering the metric cardinality. Note that the numbers usually don’t add up to 100% because cardinality is calculated as a combination of several attributes. However, it is a good indicator of what your highest contributing attributes are.

This information is only for the last day (since UTC midnight), which is when cardinality resets (both account and metric).

You can also view specific information about an attribute, which can help you make decisions about whether or not that attribute is useful. You can see a sample of the attribute values by clicking the menu on the right side of the attribute row.

Account pruning rules

A common way to keep account cardinality under its limit is to implement pruning rules for metrics and attributes you don't consider important.

View account pruning rules

You can view pruning rules by clicking into a metric and then clicking View Pruning Rules. This will bring up a scrollable list of all pruning rules associated with an account. You can also delete pruning rules associated with an account in this view:

Create a pruning rule

You can create a pruning rule based on attributes within the UI. To create a pruning rule:

1. Select the checkboxes next to the attributes you want to prune.

2. Click Create pruning rule.

3. In the pop-up with information about how the pruning rule affects the metric's cardinality, enter a rule name or description.

4. Click Submit to generate the pruning rule.

It's important to note that pruning rules are instantaneous. The information will be pruned instantly after creating the rule, so make sure you really want to prune the data.

If you create a pruning rule, you can still delete the pruning rule after it's made, but before it's deleted, the rule may prune a little bit of data.

Delete a pruning rule

To delete a pruning rule:

1. Click View All Pruning Rules.
2. Scroll to the relevant pruning rule, and then click Delete.

What's next

If you want to learn more about high cardinality, see Understand and query high cardinality metrics.