semconv.com

Specificationv1.52.0

since v1.51.0
12 breaking19 notable29 editorial

From the upstream release notes

Read the full release notes

Breaking12

Requirements added to, removed from, or restrengthened in a document marked Stable.

  • rule droppedrequirementconfiguration/sdk-environment-variables.md#durationMUST removed: Any value that represents a duration MUST be an integer representing a number of milliseconds.
  • rule droppedrequirementconfiguration/sdk-environment-variables.md#durationMUST removed: The value is non-negative - if a negative value is provided, the implementation MUST generate a warning, gracefully ignore the setting and use the default value if it is defined.
  • rule droppedrequirementconfiguration/sdk-environment-variables.md#enumMUST removed: If a null object (empty, no-op) value is acceptable, then the enum value representing it MUST be "none".
  • rule droppedrequirementconfiguration/sdk-environment-variables.md#timeoutMUST removed: Any value that represents a timeout MUST be an integer representing a number of milliseconds.
  • rule droppedrequirementconfiguration/sdk-environment-variables.md#timeoutMUST removed: The value is non-negative - if a negative value is provided, the implementation MUST generate a warning, gracefully ignore the setting and use the default value if it is defined.
  • new rulerequirementconfiguration/common.md#durationMUST added: Any value that represents a duration MUST be an integer representing a number of milliseconds.
  • new rulerequirementconfiguration/common.md#durationMUST added: The value is non-negative - if a negative value is provided, the implementation MUST generate a warning, gracefully ignore the setting and use the default value if it is defined.
  • new rulerequirementconfiguration/common.md#enumMUST added: If a null object (empty, no-op) value is acceptable, then the enum value representing it MUST be "none".
  • new rulerequirementconfiguration/common.md#numericMUST added: > The following paragraph was added after stabilization and the requirements are > thus qualified as "SHOULD" to allow implementations to avoid breaking changes. > For new > implementations, these should be treated as M...
  • new rulerequirementconfiguration/common.md#timeoutMUST added: Any value that represents a timeout MUST be an integer representing a number of milliseconds.
  • new rulerequirementconfiguration/common.md#timeoutMUST added: The value is non-negative - if a negative value is provided, the implementation MUST generate a warning, gracefully ignore the setting and use the default value if it is defined.
  • new rulerequirementtrace/sdk.md#alwaysrecordMUST added: Based on the decision from the wrapped root sampler, AlwaysRecord MUST behave as follows:

Notable19

  • removedsectionconfiguration/sdk-environment-variables.md#configuration-typesSection "Configuration types" removed.
  • removedsectionconfiguration/sdk-environment-variables.md#durationSection "Duration" removed.
  • removedsectionconfiguration/sdk-environment-variables.md#integerSection "Integer" removed.
  • removedsectionconfiguration/sdk-environment-variables.md#timeoutSection "Timeout" removed.
  • strengthrequirementmetrics/sdk.md#instrument-enabledMUST relaxed to SHOULD: Otherwise, it SHOULD return true.
  • stabilitydocumentsdk_exporters/zipkin.mdStatus changed from Stable to Deprecated.
  • new rulerequirementconfiguration/common.md#enumMAY added: For configuration sources which accept a known value out of a set, i.e., an enum value, implementations MAY support additional values not listed here.
  • new rulerequirementconfiguration/common.md#enumSHOULD added: When reporting configuration errors, implementations SHOULD display the original user-provided value to aid debugging.
  • new rulerequirementconfiguration/common.md#integerSHOULD added: If an implementation chooses to support an integer-valued configuration source, it SHOULD support non-negative values between 0 and 2³¹ − 1 (inclusive).
  • new rulerequirementconfiguration/common.md#integerMAY added: Individual SDKs MAY choose to support a larger range of values.
  • new rulerequirementconfiguration/common.md#numericSHOULD added: For sources accepting a numeric value, if the user provides a value which is outside the valid range for the configuration item, the implementation SHOULD generate a warning and gracefully ignore the setting, i.e., trea...
  • new rulerequirementconfiguration/common.md#numericSHOULD NOT added: In particular, implementations SHOULD NOT assign a custom interpretation e.g. to negative values if a negative value does not naturally apply to a configuration and does not have an explicitly specified meaning, but tre...
  • new rulerequirementconfiguration/common.md#opentelemetry-common-configuration-specificationMAY added: Implementations MAY choose to allow configuration via any configuration source in this specification, but are not required to.
  • new rulerequirementconfiguration/common.md#opentelemetry-common-configuration-specificationSHOULD added: If they do, they SHOULD follow the guidance listed in this document.
  • new rulerequirementconfiguration/common.md#timeoutSHOULD added: Implementations SHOULD interpret timeout zero values (i.e. 0 indicating 0 milliseconds) as no limit (i.e. infinite).
  • new rulerequirementconfiguration/common.md#timeoutMAY added: In practice, implementations MAY treat no limit as "a very long time" and substitute a very large duration ( e.g. the maximum milliseconds representable by a 32-bit integer).
  • new rulerequirementconfiguration/sdk-environment-variables.md#opentelemetry-environment-variable-specificationSHOULD added: They SHOULD also follow the common configuration specification.
  • new rulerequirementmetrics/sdk.md#instrument-enabledMUST added: The instrument Enabled MUST return false when either:
  • new rulerequirementsdk_exporters/zipkin.md#opentelemetry-to-zipkin-transformationMUST added: Existing stable Zipkin exporters MUST > continue to be supported for at least one year after the artifact is deprecated, > following the SDK stability guarantees. > Implementing a Zipkin exporter is not required for new...

Editorial29