# Datadog APM Span Ingestion and Indexing Beyond Host Allotments Canonical: https://www.pointfive.co/efficiency-hub/inefficiencies/datadog-apm-span-ingestion-and-indexing-beyond-host-allotments Datadog APM bills per APM host, and each host brings a monthly allotment of ingested span volume and indexed spans. By: PointFive Updated: 2026-09-28 [Cloud Efficiency Hub](https://www.pointfive.co/efficiency-hub) The short version Datadog APM bills per APM host, and each host brings a monthly allotment of ingested span volume and indexed spans. PointFive Research Cloud cost research at PointFive Datadog service [Datadog APM](https://www.pointfive.co/efficiency-hub/cloud-services/datadog-apm) Category [Other](https://www.pointfive.co/efficiency-hub/service-category/other) Reference CER-0542 Type Excessive Ingestion or Processing ## Explanation Why the waste happens and who it affects. Usage above the pooled allotment is billed separately per GB ingested and per million spans indexed. Overage usually comes from a few high-throughput services: a sampling rule or DD\_TRACE\_SAMPLE\_RATE set to 100% during an investigation and never reverted, a root service whose sampling decision pulls in every downstream span of the trace, or custom retention filters that index a large share of all spans at a high retention rate. Much of this data is not needed. Datadog computes request, error and duration (RED) trace metrics from 100% of application traffic regardless of sampling, and the default Intelligent Retention Filter keeps a representative set of traces without counting toward indexed span usage. Ingesting and indexing far beyond that pays for spans that are rarely searched, and because ingestion and indexing are metered separately from the host fee, the overage often goes unnoticed until the invoice. ## Billing model The pricing dimensions that drive this cost. APM combines a per-host fee with two volume dimensions that are only charged above the allotment pooled across all APM hosts. APM host Billed per underlying host per month on the 99th percentile of hourly counts ($31 per host per month on the published APM billing page), including 150 GB of ingested spans and 1 million indexed spans Ingested spans Billed per GB of span data ingested per month above the pooled allotment ($0.10 per GB on the published APM billing page) Indexed spans Billed per million spans indexed by retention filters per month above the pooled allotment ($1.70 per million with default 15-day retention) Intelligent Retention Filter Diversity sampling and 1% flat sampling whose indexed spans do not count toward indexed span usage ## How to detect 5 checks to find it in your estate. - On the APM Ingestion Control page, check the Monthly Ingestion panel: it shows ingestion so far this month, projected end-of-month ingestion and the monthly allotment, and flags when the projection exceeds the allotment - Sort the Ingestion Control service table by Downstream Bytes/s to find the root services whose sampling decisions drive most ingested volume, and check the Configuration column for APM Local or APM Remote rules that set a high fixed rate - Chart datadog.estimated\_usage.apm.ingested\_bytes by service and env, and datadog.estimated\_usage.apm.indexed\_spans by retention filter, against the allotment (150 GB and 1 million indexed spans per APM host) - On the Retention Filters settings page, review the Retention Rate and Spans Indexed columns for custom filters that index broad queries at or near 100% - Open the service ingestion summary for the top services and check the Ingestion reasons breakdown and the Sampling rates by resource table for rules at or near 100%, such as a global DD\_TRACE\_SAMPLE\_RATE of 1.0 left from a debugging session ## How to fix 5 ways to remove the waste. - Move high-throughput services to adaptive sampling with a monthly ingestion target, or set resource-based sampling rates so health checks and high-volume endpoints are sampled lower than business-critical ones - Lower the Agent head-based target with DD\_APM\_TARGET\_TPS (10 traces per second per Agent by default), reduce DD\_APM\_ERROR\_TPS, or keep the rare sampler disabled when the default mechanism drives most volume - Narrow custom retention filters to the queries that matter, such as errors, high latency and key endpoints, and lower their retention rates; rely on the Intelligent Retention Filter for general sampling - Build dashboards and monitors on trace.\* metrics, which are computed from 100% of traffic, or on metrics generated from spans, instead of indexing spans only to count them; span-based metrics are billed as custom metrics, so avoid high-cardinality group-bys - Tradeoff: reducing ingestion lowers the number of traces available for investigation and affects count-based metrics from spans and trace analytics monitors, so check those before lowering rates ## Documentation Vendor references for pricing and configuration. - [APM Billing docs.datadoghq.com](https://docs.datadoghq.com/account_management/billing/apm_tracing_profiler/) - [Ingestion Controls docs.datadoghq.com](https://docs.datadoghq.com/tracing/trace_pipeline/ingestion_controls/) - [Ingestion volume control with APM Distributed Tracing docs.datadoghq.com](https://docs.datadoghq.com/tracing/guide/trace_ingestion_volume_control/) - [Trace Retention docs.datadoghq.com](https://docs.datadoghq.com/tracing/trace_pipeline/trace_retention/) - [Generate Custom Metrics from Spans and Traces docs.datadoghq.com](https://docs.datadoghq.com/tracing/trace_pipeline/generate_metrics/) ## Related inefficiencies [Browse the library](https://www.pointfive.co/efficiency-hub) - Datadog Log Management CER-0539 ### [Indexing All Ingested Datadog Logs Without Exclusion Filters](https://www.pointfive.co/efficiency-hub/inefficiencies/indexing-all-ingested-datadog-logs-without-exclusion-filters) Datadog Log Management charges separately for ingesting logs and for indexing them. At list prices, indexing a million 1 KB events for 15 days ($1.70) costs many times more than ingesting that gigabyte ($0.10). By default a log index has... Other - Datadog Custom Metrics CER-0540 ### [High-Cardinality Tags Inflating Datadog Custom Metrics](https://www.pointfive.co/efficiency-hub/inefficiencies/high-cardinality-tags-inflating-datadog-custom-metrics) Under Datadog's cardinality-based pricing, a custom metric is a unique combination of a metric name and tag values, including the host tag. Every new tag value creates a new billable timeseries, so a metric tagged with an unbounded... Other - Datadog Infrastructure Monitoring CER-0544 ### [Datadog Container Usage Exceeding Per-Host Allotment](https://www.pointfive.co/efficiency-hub/inefficiencies/datadog-container-usage-exceeding-per-host-allotment) Each Datadog Infrastructure host license on the Pro or Enterprise plan includes a small number of monitored containers, 5 or 10 per host depending on the plan, pooled across the whole account. Dense Kubernetes nodes, sidecar-heavy pods,... Other --- Source: the public page above. Product screenshots and illustrative interfaces are examples, not live customer data.