APM Billing

APM Billing

APM & Continuous Profiler powers you to find service bottlenecks, analyze distributed traces and code performance across your microservices architecture.

There are two options available for pricing, depending on whether APM and Profiling are bundled. Additionally, using the Tracing Without Limits feature with APM allows you to slice and dice your application data with spans indexed by tagged-based custom retention filters.

Billing ParameterPriceIngested and Indexed SpansBilling
APM Host$31 per underlying APM host per month1 million Indexed Spans and 150 GB of Ingested Spans included per month with every APM host.Datadog records the number of APM hosts you are concurrently monitoring in the Datadog APM service once an hour. On a high watermark plan (HWMP), these hourly measurements are ordered from highest to lowest at the end of the month, and Datadog charges based on the eighth highest measurement. More information.
APM & Continuous Profiler$40 per underlying APM host and includes the Continuous Profiler with 4 profiled containers per month.Same as APM HostDatadog records the number of unique Continuous Profiler hosts you are concurrently monitoring with the Datadog Continuous Profiler service once per hour. The hourly measurements and billing are conducted the same as for APM Hosts.
Fargate$2 per concurrent task per month65,000 Indexed Spans and 10 GB of Ingested Spans included in pricing.Datadog records the number of task instances you are monitoring in the Datadog APM service at five-minute intervals. Datadog aggregates the interval-based measurements at the end of the month and charges you based on the total number of hours your applications were run and monitored. More information.
Indexed span$1.70 per million Indexed Spans per monthBilled when usage is in excess of Indexed Spans included with every APM hostAn Indexed span is an individual request against an individual service in your stack. Datadog charges based on the total number of spans indexed with retention filters or legacy Analyzed Spans to the Datadog APM service at the end of the month. More information.
Ingested span$.10 per GB Ingested Spans per monthBilled when usage is in excess of Ingested Spans included with every APM hostAn Ingested span is an individual request against an individual service in your stack. Datadog charges based on the total number of gigabytes of spans ingested to Datadog at the end of the month. More information.

Note: If you’re using a container based environment, you get billed for the underlying host deploying the Datadog Agent.

Note: One profiled container is a container that is running the Continuous Profiler service. This does not include containers that are not being profiled. For instance, a DNS service container that is NOT profiled, running concurrently with your application container that IS profiled, is not counted towards the four profiler containers allotment.

For more information, see the Pricing page.

Deployment scenarios

Sample cases illustrate annual billing rates with default 15 days Indexed Span retention. Contact Sales or your Customer Success Manager to discuss volume discounts for your account.

APM Hosts, indexed spans, and extra ingested spans without profiler

Using 5 hosts and sending 30 million Indexed Spans, with 900 GB of total Ingested Spans without Profiler.

Billable UnitQuantityPriceFormulaSubtotal
APM Hosts5$31 per host5 * $31$155
Indexed Spans30 million5 million included with 5 APM hosts. $1.70 per million for additional 25 million Indexed Spans25 * $1.70$42.50
Ingested Spans900 GB750 GB included with 5 APM hosts. $.10 per GB for additional 150 GB of Ingested Spans.150 * $.10$15
Total$155 + $42.50 + $15$212.50 per month

APM and Profiler with six profiled containers per host

APM & Continuous Profiler for five hosts with six apps running in separate containers per each host.

Billable UnitQuantityPriceFormulaSubtotal
APM & Continuous Profiler5$40 per host5 * $40$200
Profiled containers6 per host$2 per additional container per host. In this case there are 6 - 4 = 2 additional containers for each host2 * $2 * 5 hosts$20
Total$200 + $20$220 per month

APM Hosts, Fargate, and indexed spans without profiler

Using 5 hosts, sending 20 million Indexed Spans, and have deployed APM on average 20 Fargate Tasks over the month.

Billable UnitQuantityPriceFormulaSubtotal
APM Hosts5$31 per host5 * $31$155
Fargate Tasks20$2 per task20 * $2$40
Indexed Spans20 million5 million included with 5 APM hosts. 1.3 million included with 20 Fargate tasks. $1.70 per million for additional 13.7 million Indexed Spans13.7 * $1.70$23.29
Total$155 + $40 + $23.29$218.29 per month

APM & Continuous Profiler, services, containers, and indexed spans

APM & Continuous Profiler for service 1 running on container 1, service 2 running on container 2. Both Containers are running on 1 host and are sending 20 million Indexed Spans on App Analytics.

Billable UnitQuantityPriceFormulaSubtotal
APM & Continuous Profiler1$40 per host1 * $40$40
Profiled containers2$0 as profiled containers are within the 4 per APM Host allotment.
Indexed Spans20 million1 million included with 1 APM host. $1.70 per million for additional 19 million Indexed Spans19 * $1.70$32.30
Total$40 + $32.30$72.30 per month

APM hosts with dynamic scaling, containers, Fargate, and no indexed spans without profiler

App 1 running on 20-40 containers which are deployed on 4-8 host instances, app 2 running on 10-30 Fargate tasks. Assuming, the 99th percentile usage of EC2 instances is 7, and average of Fargate Tasks over the month is 28.

Billable UnitQuantityPriceFormulaSubtotal
APM Hosts7$31 per host7 * $31$217
Fargate Tasks28$2 per task28 * $2$56
Total$217 + $56$273 per month

Note: The container count does not matter if the deployed Agent is on the EC2 instances.

APM and Continuous Profiler with Kubernetes nodes and indexed spans

APM & Continuous Profiler for apps with a Datadog Agent running on 20 worker nodes in Kubernetes sending 20 million Indexed Spans. 10 of these worker nodes have 8 pods each with 1 container per pod, the other 10 have 2 pods each with 1 container per pod.

Billable UnitQuantityPriceFormulaSubtotal
APM & Continuous Profiler (Nodes)20$40 per host20 * $40$800
Profiled containers100 in aggregate$2 per additional container. In this case 20 hosts would allow up to 80 containers but there are 20 containers summed across two hosts: 100-80 = 20 additional containers$2 * 20 hosts$40
Indexed Spans20 million20 million included with 20 APM hosts (nodes). No additional Indexed Spans0 * $1.700
Total$800 + $40$840 per month

For Kubernetes, APM & Continuous Profiler are priced by nodes not by pods.

Lambda functions and indexed spans

An AWS-Lambda based serverless application being invoked 10 million times in a month while sending 10 million Indexed Spans.

Billable UnitQuantityPriceFormulaSubtotal
Lambda Function Invocations10 million$5 per month10 * $5$50
Indexed Spans10 million150,000 Indexed Spans included with each 1 million Lambda invocations. $1.70 per million additional Indexed Spans8.5 * $1.70$14.45
Total$50 + $14.45$64.45 per month

FAQ

1. What is classified as an APM host for billing?

A host is a physical or virtual operating system instance. Datadog records the number of hosts you are concurrently monitoring in the Datadog Infrastructure service once an hour. For billing APM, number of hosts with APM installed and sending traces are calculated every hour. At the end of the month, you are billed based on your 99th percentile usage for APM hosts.

2. How is billing calculated when deploying one Agent per container?

It is recommended to setup running one agent per underlying host for container deployment. If you still choose to run one agent per container, then each container is treated as a single host. The price is then (Price Per APM host) * (Number of containers)

3. What is classified as an APM Fargate task for billing?

A Fargate task is a collection of containers that are scheduled to run on AWS Fargate as a serverless compute engine. Datadog records the number of tasks you are concurrently monitoring in Datadog at five-minute intervals. For billing APM, Datadog bills based on the average number of Fargate tasks that send traces to Datadog per hour across the month of your account.

4. What happens to your bill when scaling your environment?

Your APM bill is calculated using the top 99 percentile of active agents sending traces every hour of each month. At the end of the month, Datadog disregards the top 1% value, giving a shield against being billed for unexpected spikes.

5. Do you get charged for pause containers in Kubernetes?

Kubernetes creates pause containers to acquire the respective pod’s IP address and set up the network namespace for all other containers that join that pod. Datadog excludes all pause containers from your quota and does not charge for them (requires Agent 5.8+). For Kubernetes, APM is priced by nodes not by pods.

6. How is the host billing related to your services?

APM is billed on the basis of hosts deployed with agents sending traces and not services. Tracing without Limits is billed on the basis of Indexed and Ingested span count. To estimate how many ingested and indexed spans each of your services is sending, see the Usage Metrics documentation.

7. What happens to your existing App Analytics filters?

As of October 20, 2020, all existing App Analytics filters are automatically transitioned to Retention Filters. You can continue to let the filters remain unchanged or modify them as needed. Transitioned filters are marked with an i representing Legacy App Analytics Filters within the retention filters page.

8. How do you estimate your ingested or indexed span volume?

Datadog provides the metrics datadog.estimated_usage.apm.ingested_bytes and datadog.estimated_usage.apm.ingested_spans for monitoring ingested and indexed span volume. More information is available in the Usage Metrics documentation.

9. Is the Continuous Profiler available as a standalone product?

Yes. Let Datadog know if you are interested in buying the Continuous Profiler without APM. Reach out to Sales or your Customer Success Manager.

Further Reading

Additional helpful documentation, links, and articles: