View and manage data retention

Using the Data retention UI page, review and manage how long your data will be stored in the New Relic database (NRDB). Keep data for shorter periods to protect sensitive information or for longer periods to notice trends over time. At a glance, get the sources of data that you’re storing, when they were last updated, and the length the data is stored. Retention settings range from 8 to 395 days depending on the data's source and namespace classifications in New Relic.

Find the data retention UI

The Data retention tab is located in the Data management hub.

Retention periods

The following table shows the default retention settings for our New Relic One pricing model.

Adjust retention values

You can adjust your retention periods to better support your data needs. For example, you might reduce your contracted retention values to optimize query performance. Conversely, you might increase retention for some data types to support more granular analysis over longer periods of time, or to meet data compliance requirements.

For each additional month (30 days) of retention, the cost is $0.05 per GB ingested per month.

Requirements for adjusting retention

Pricing model-related requirements:

• New Relic One pricing model: requires Pro or Enterprise edition to edit retention.
• Original pricing model: requires Insights Pro to edit retention.

The ability to edit data retention is impacted by which user model a user is on:

• New Relic One user model: requires the Billing user role.
• The original user model: requires role of Owner, Admin, or the Data retention manager add-on role.

Reduce retention values

When you reduce the number of days that specific data sources are retained, existing data of that source type won't immediately disappear. That data might be visible and returned in queries for up to several weeks. We have safeguards in place to ensure you don't accidentally delete data, but this also means that reducing retention values isn't a good strategy for quickly removing data.

That said, your change is registered in the system within 12 hours and new data of the adjusted source type will be retained for the new duration.

Note that reducing data retention below your contracted retention values will not reduce any New Relic data ingest charges. Conversely, increasing retention for targeted telemetry data types is likely to trigger a conversation with New Relic about adjusting your data ingest charges.

Copy retention values

If you have multiple accounts in your organization, you can also copy retention values from one account to other accounts. To do this, select the Copy settings to other accounts link, and then select the accounts you want to apply the retention values to.

Deleting data

Once telemetry data (events, metrics, logs, traces) is reported to New Relic and available for querying, that data cannot be edited or deleted. This is a purposeful design decision that optimizes New Relic's speed and performance. Data will expire after its data retention period ends. If you need to delete data that has already been reported, contact your New Relic representative.

Manage retention via API

You can change your retention settings with NerdGraph as well as the UI. For the following examples, provide any necessary information, and run the code in the NerdGraph GraphiQL explorer.

Your retention changes take effect within 24 hours after updating.

query {
 actor {
   account(id: YOUR_ACCOUNT_ID) {
     dataManagement {
       customizableRetention {
         eventNamespaces {
           namespace
         }
       }
     }
   }
 }
}

query {
 actor {
   account(id: YOUR_ACCOUNT_ID) {
     dataManagement {
       eventRetentionRules {
         id
         deletedAt
         deletedById
         createdAt
         createdById
         retentionInDays
         namespace
       }
     }
   }
 }
}

query {
 actor {
   account(id: YOUR_ACCOUNT_ID) {
     dataManagement {
       eventRetentionRule(namespace: "APM") {
         id
         deletedById
         deletedAt
         createdById
         createdAt
         retentionInDays
         namespace
       }
     }
   }
 }
}

mutation {
 dataManagementCreateEventRetentionRule(accountId: YOUR_ACCOUNT_ID, namespace: "APM", retentionInDays: 8) {
   id
   deletedById
   deletedAt
   createdById
   createdAt
   retentionInDays
   namespace
 }
}

mutation {
 dataManagementDeleteEventRetentionRule(accountId: YOUR_ACCOUNT_ID, namespace: "APM") {
   id
   deletedById
   deletedAt
   createdById
   createdAt
   retentionInDays
   namespace
 }
}

mutation {
  dataManagementCopyRetentions(
    sourceAccountId: SOURCE_ACCOUNT_ID,
    destinationAccountIds: [DESTINATION_ACCOUNT_ID_1, DESTINATION_ACCOUNT_ID_2, DESTINATION_ACCOUNT_ID_3])
  {
    success
    failure
  }
}

Metric timeslice data

Metric timeslice data is generated by APM (both custom metrics and default metrics), mobile, and browser agents. A metric timeslice consists of three parts: a metric name, the segment of time the metric represents (the "timeslice"), and a numeric value (the measurement). APM, browser, and mobile agents can collect thousands of metric timeslices per minute for a variety of performance metrics, such as error rate, bandwidth usage, and garbage collection time.

Metric timeslice data is aggregated like this:

• first to 1-minute rollups that are stored for 8 days
• the 1-minute rollups are also added to 1-hour roll-ups that are retained for 90 days

Key metric data is stored indefinitely. We meter on the size of the initial 1-minute rollup, and not on any of the subsequent rollups. Read on to learn the key metrics by product.

Here are key metric timeslice data that is stored indefinitely:

Raw dimensional metric data

Raw dimensional metric data is handled differently than is metric timeslice data. Raw sub-minute dimensional metric data is stored for 30 days. Beyond 30 days, 1-minute resolution roll-up events are created and retained for 13 months. As with metric timeslice data, we meter and bill on the size of the raw ingest, not the subsequent rollups.

Metric retention periods are not editable.

Infrastructure data

Infrastructure monitoring downsamples your data as it's generated. All infrastructure monitoring metric data types (including metrics from on-host integrations) will display different granularity depending on the age of the data and the size of the time window. The following table illustrates when different downsampled buckets will be used, both in the infrastructure UI and for queries:

Account inheritance

The data retention settings of a parent account are inherited by its children accounts. In this context, an organization's first account is treated as the parent account, and other accounts added to the organization inherit the settings in that account. You also have the option to override those settings and set account-specific data retention settings.