Latest Dynatrace
With Metrics powered by Grail, you can experience the full power of Grail with limitless ingest and configurable retention.
Metrics powered by Grail
Metrics Classic
Query language
DQL
Metric selector
Ingest limits
Limitless (excluding highly volatile dimensions)
Query limits
500 million data points
20 million data points
Retention
15 months included by default, with option to increase to 10 years
5 years, not configurable
Metrics powered by Grail allows you to monitor highly dynamic environments that exceed Metrics Classic limits, for example, metrics with overall cardinality above 1 million. However, this does not include highly volatile metrics that can negatively impact metrics storage, performance, and costs. Using a timestamp or short-lived or unique IDs are all examples of such misconfiguration. Dynatrace therefore may restrict or reject metric configurations that don't follow best practices.
Other than enforcing reserved field names, the Metrics API v2 and Metric ingestion protocol enforce the same usage and limits for DPS and non-DPS licensed tenants.
As a DQL field, your metric.key should align with DQL field naming rules.
Examples of valid metric.key values are:
http_request_durationhost.cpu.temperature`my-custom-metric` will require backticks in your DQL query, as hyphens (-) are special characters.Grail reserves the dt. prefix for Dynatrace-controlled metrics, such as dt.host.cpu.usage. Metric data ingested with the dt. prefix will be dropped.
Grail reserves Dynatrace system fields and other field names with the prefix dt.. Dimensions ingested with the dt. prefix are ignored.
You can learn more about the reserved fields in the Semantic Dictionary.