Account usage for CU-based APM and Infrastructure

This document explains the New Relic Account usage UI page for accounts that have compute unit (CU) pricing. This information applies to:

  • All New Relic Infrastructure accounts
  • New Relic APM accounts with CU-based pricing

If you have APM with host-based pricing, New Relic Browser, or New Relic Mobile, see Account usage UI page (non-CU).

Understand the usage UI

To view the Account usage page for CU-based accounts: From the New Relic UI's account dropdown, select > Account settings > APM usage or Infrastructure usage.

(account dropdown) > Account settings > Usage: Use the Account usage UI page to understand your usage of New Relic Infrastructure and APM accounts with CU-based pricing. For New Relic Browser, Mobile, and host-based APM, see Account usage (non-CU).

Here are ways to gain more detail about your compute unit usage:

  • Average CUs per month are displayed in the upper right of the usage chart.
  • To view a specific month's usage: mouse over any bar in the chart to view CUs used.
  • The Usage details table provides the CU usage for your account broken down by individual hosts. The Usage (CUs) column displays the total number of CUs used for each host. You can re-order the table by selecting column headers. The table provides data for the last three days UTC, up to the latest complete hour.
  • For details on the active host data, download a comma-separated value (CSV) file by selecting download Download active hosts' usage as CSV.
  • For APM only: Use the CU usage API to obtain and track your APM usage.

All measurements of time in the UI use Coordinated Universal Time (UTC). For example, if the table displays "since Nov 19," it means "since Nov 19th, 12:00am UTC."

If you use Docker or other Linux container environments, it's recommended you upgrade to the latest New Relic APM agent. For more information, see Container issues.

Account usage terminology

Here are definitions for some of the table column headers and terms used on this UI page:

Term Definition
Host ID

Host ID is the unique identifier for that host. If the host is in AWS, we use the AWS-provided instance ID. For other hosts, New Relic assigns a host ID.

Host provider

Host provider is the name of the cloud provider that hosts the host instances connected to New Relic. This value may be blank for some non-AWS cloud providers or if your hosts are located in your own datacenter.

Instance An instance refers to a host (either a physical machine, a virtual machine, or a cloud-based instance) associated with a single operating system. In this context, an instance is another word for host.
Instance type Instance type is the type of AWS host connected to New Relic: for example, t2.micro or m4.large. The value will be blank for non-AWS cloud providers, or if the host is in your own datacenter.
Instance size

Host instance size is based on the computing power and memory used by the host. The host instance size is calculated as the sum of logical CPU cores and total gigabytes of RAM for a given host. For example, a host with 2 CPU cores and 2GB RAM has a size of 4.

The maximum host instance size is capped at 16, which consumes 12000 CUs for that month.

Hours used The total number of hours that the host is connected to New Relic for that month. For a host connected for a full month, the maximum value is 750 hours.
Usage (CUs) Usage (CUs) indicates the total number of compute units used for that month, broken out by hosts. A compute unit (CU) is the unit of measurement for New Relic's CU-based pricing model. For more on how CUs work, see CU-based pricing.
Logical CPU core A logical CPU core is a processing unit is capable of executing its own thread in parallel with other logical cores. A physical core can have multiple logical cores.

The downloadable CSV file contains the active host data for the past two days prior to being downloaded. The CSV file may contain the column headers listed below in addition to the UI column headers listed above. The following table contains headers from both APM and Infrastructure files.

CSV file header Definition
Total RAM Gigabytes of RAM available to the host.
Logical processors The number of logical CPU cores available to the host.
Docker container ID

The unique ID of the Docker container. If the host is running multiple containers with different IDs, New Relic will use the first one detected.

Cloud/AWS availability zone The availability zone of the AWS/cloud host instance.
Cloud/AWS instance ID The vendor-provided cloud instance (host) ID.
Account ID New Relic account ID for the host.
Business rule

APM only (not Infrastructure): The business rule provides information about how New Relic APM reports the host and its instance size. This will help you better understand your data and perhaps lead to a New Relic agent upgrade. Business rules include:

  • cloud_provider_data: The instance size is calculated based on cloud instance type. The cloud provider's instance ID is used as the Host ID; for example, an AWS instance ID of i-12345678.
  • agent_collected_calculated_data: The agent on this host reported CPU and RAM measurements that were used to calculate the instance size.
  • update_agent_for_container_environment: This host is running an agent that does not report enough information to identify the host a container is running in. For more about this issue, see container/Docker issues.
  • old_agent_data_cap: This host is running an agent that does not report any cloud provider identification, or any RAM or CPU information. The instance size is therefore calculated as 16. Upgrade the agent on this host to get more accurate instance size reporting.
  • missing_cpu_data_cap: This host is running an agent that did not report a number of logical processors. The instance size is therefore calculated as 16.
  • missing_ram_data_cap: This host is running an agent that did not report an amount of available RAM. The instance size is therefore calculated as 16.
  • missing_cpu_and_ram_data_cap: This host is running an agent that did not report a number of logical processors or amount of available RAM. The instance size is therefore calculated as 16.
  • max_instance_size_cap: The calculated instance size for the host is greater than the maximum size, so it has been capped at 16.

Sub-accounts and usage

You can view usage data from your master account, or master accounts and sub-accounts combined. To switch between master account usage and master + sub-account usage, select filterFilter by account.

Troubleshooting and UI error messages

If there is an account usage issue detected by New Relic, you may see a message in the UI. Messages include:

Some of your hosts are running containers: Update your agent to get accurate usage information.

Some New Relic APM agent versions may miscount containers as hosts, which may lead to over-reporting of compute unit (CU) usage. Some newer APM agent versions have fixed this problem for Linux-based containers, including Docker. For more on this issue, see Container issues.

This is not an issue for New Relic Infrastructure.

Default instance size applied: Update your agent to report accurate instance size.

If your APM agent version is outdated, New Relic is unable to determine the actual host size, and the maximum size of 16 will be applied be default. To accurately report usage information, upgrade to the latest APM agent version. Account usage data requires:

  • Go: 1.0 or higher
  • Java: 3.18.0 or higher
  • .NET: or higher
  • Node.js: 1.21.0 or higher
  • PHP: 5.5.0 or higher
  • Python: or higher
  • Ruby: or higher

This is associated with the business rule old_agent_data_cap that may be found in the downloadable CSV file.

Default instance size applied: System was not able to detect instance size

As a result of some agent issues or an unsupported environment, New Relic cannot accurately identify the size of the host. As a result, the default instance size of 16 will be applied.

Custom instance size applied: Instance size obtained from customer provided information

The host instance size was obtained from customer information, and a custom instance size has been applied as a result.

APM issue with containers (including Docker)

Some New Relic APM agents may miscount containers as hosts, which may lead to over-reporting of compute unit (CU) usage. New Relic Infrastructure does not have this issue.

To fix this issue for Linux-based containers (including Docker), upgrade to these APM agent versions:

  • Java: 3.42.0 or higher
  • Node.js: 2.1.0 or higher
  • PHP: or higher
  • Python: or higher
  • Ruby: or higher
  • Go agent: 1.11 or higher
  • .NET agent: 6.19.330.0 or higher

For more help