azure-servicebus_7.14.2

7.14.2 (2025-04-09)

Bugs Fixed

• Fixed a bug where max number of messages was not being requested when receiving from the service due to an incorrect link credit calculation. (#40156)

Other Changes

• Added support for handling a C# DateTime.MinValue timestamp, which is returned by the service as a sentinel for time which is not set.
• Deprecating uamqp_transport in favor of pyAMQP transport. The uamqp_transport will be removed in the next minor release.
• Fixed aiohttp websocket library showing a deprecation warning due to an incorrect timeout type (#40429)
• Dropped support for Python 3.8

azure-mgmt-iothub_4.0.0

4.0.0 (2025-04-09)

Breaking Changes

• Removed subfolders of some unused Api-Versions for smaller package size. If your application requires a specific and non-latest Api-Version, it's recommended to pin this package to the previous released version; If your application always only use latest Api-Version, please ignore this change.

azure-ai-ml_1.26.2

1.26.2 (2025-04-08)

Bugs Fixed

• Made AI Search connections property optional while creating capability host.

azure-monitor-opentelemetry_1.6.6

1.6.6 (2025-04-07)

Features Added

• Add logging_format as configuration option in api
  (#40057)
• Enable Azure AI Agents instrumentation
  (#40043)

azure-monitor-opentelemetry-exporter_1.0.0b36

1.0.0b36 (2025-04-07)

Features Added

• Support syntheticSource from user_agent.synthetic.type semantic convention
  (#40004)
• Support server.address attributes when converting Azure SDK messaging spans to envelopes
  (#40059)
• Update AKS check to use KUBERNETES_SERVICE_HOST
  (#39941)
• Enabled Entra ID Credential configuration via env var
  (#40237)

azure-batch_15.0.0b2

15.0.0b2 (2025-03-01)

Features Added

• Force delete/terminate job or job schedule:

  • Added force parameter of type Boolean to delete_job, terminate_job, delete_job_schedule, and terminate_job_schedule

• Support for compute node start/deallocate operations:

  • Added start_node, deallocate_node methods to BatchClient and AsyncBatchClient

• Container task data mount isolation:

  • Added containerHostBatchBindMounts of type List<ContainerHostBatchBindMountEntry> to BatchTaskContainerSettings

• Patch improvements for pool and job:

  • Added displayName, vmSize, taskSlotsPerNode, taskSchedulingPolicy, enableInterNodeCommunication, virtualMachineConfiguration, networkConfiguration, userAccounts, mountConfiguration, upgradePolicy, and resourceTags to BatchPoolUpdateContent
  • Added networkConfiguration to BatchJobUpdateContent

• Confidential VM support:

  • Added confidentialVM to SecurityTypes.
  • Added securityProfile of type VMDiskSecurityProfile to ManagedDisk

• Support for shared and community gallery images:

  • Added sharedGalleryImageId and communityGalleryImageId to ImageReference

• Re-add support for BatchCertificate (temporary since this feature is deprecated):

  • Added create_certificate, list_certificates, cancel_certificate_deletion, delete_certificate, andget_certificate methods to BatchClient and AsyncBatchClient

Breaking Changes

• Removed get_remote_desktop method from BatchClient. Use get_node_remote_login_settings instead to remotely login to a compute node
• Removed CloudServiceConfiguration from pool models and operations. Use VirtualMachineConfiguration when creating pools
• Removed ApplicationLicenses from pool models and operations

azure-ai-evaluation_1.5.0

1.5.0 (2025-04-04)

Features Added

• New RedTeam agent functionality to assess the safety and resilience of AI systems against adversarial prompt attacks

azure-core_1.33.0

1.33.0 (2025-04-03)

Features Added

• Added native OpenTelemetry tracing to Azure Core which enables users to use OpenTelemetry to trace Azure SDK operations without needing to install a plugin. #39563
  • To enable native OpenTelemetry tracing, users need to:
    1. Have opentelemetry-api installed.
    2. Ensure that settings.tracing_implementation is not set.
    3. Ensure that settings.tracing_enabled is set to True.
  • If setting.tracing_implementation is set, the tracing plugin will be used instead of the native tracing.
  • If settings.tracing_enabled is set to False, tracing will be disabled.
  • The OpenTelemetryTracer class was added to the azure.core.tracing.opentelemetry module. This is a wrapper around the OpenTelemetry tracer that is used to create spans for Azure SDK operations.
  • Added a get_tracer method to the new azure.core.instrumentation module. This method returns an instance of the OpenTelemetryTracer class if OpenTelemetry is available.
  • A TracingOptions TypedDict class was added to define the options that SDK users can use to configure tracing per-operation. These options include the ability to enable or disable tracing and set additional attributes on spans.
    • Example usage: client.method(tracing_options={"enabled": True, "attributes": {"foo": "bar"}})
  • The DistributedTracingPolicy and distributed_trace/distributed_trace_async decorators now uses the OpenTelemetry tracer if it is available and native tracing is enabled.
    • SDK clients can define an _instrumentation_config class variable to configure the OpenTelemetry tracer used in method span creation. Possible configuration options are library_name, library_version, schema_url, and attributes.
    • DistributedTracingPolicy now accepts a instrumentation_config keyword argument to configure the OpenTelemetry tracer used in HTTP span creation.

Breaking Changes

• Removed automatic tracing enablement for the OpenTelemetry plugin if opentelemetry was imported. To enable tracing with the plugin, please import azure.core.settings.settings and set settings.tracing_implementation to "opentelemetry". #39563
• In DistributedTracingPolicy, the default span name is now just the HTTP method (e.g., "GET", "POST") and no longer includes the URL path. This change was made to converge with the OpenTelemetry HTTP semantic conventions. The full URL is still included in the span attributes.
• Renamed span attributes in DistributedTracingPolicy:
  • "x-ms-client-request-id" is now "az.client_request_id"
  • "x-ms-request-id" is now "az.service_request_id"

Bugs Fixed

• Fixed an issue where the traceparent header was not being set correctly in the DistributedTracingPolicy. The traceparent header will now set based on the context of the HTTP client span. #40074

Other Changes

• Added opentelemetry-api as an optional dependency for tracing. This can be installed with pip install azure-core[tracing]. #39563

azure-mgmt-search_9.2.0b3

9.2.0b3 (2025-04-02)

Features Added

• Client SearchManagementClient added operation group offerings
• Model SearchService added property system_data
• Model SearchService added property endpoint
• Model SearchService added property compute_type
• Model SearchService added property upgrade_available
• Model SearchService added property service_upgrade_date
• Model SearchServiceUpdate added property system_data
• Model SearchServiceUpdate added property endpoint
• Model SearchServiceUpdate added property compute_type
• Model SearchServiceUpdate added property upgrade_available
• Model SearchServiceUpdate added property service_upgrade_date
• Added enum ComputeType
• Added enum CreatedByType
• Added enum FeatureName
• Added model FeatureOffering
• Added model OfferingsByRegion
• Added model OfferingsListResult
• Added model SkuOffering
• Added model SkuOfferingLimits
• Added model SystemData
• Operation group ServicesOperations added method begin_upgrade
• Added operation group OfferingsOperations

azure-ai-evaluation_1.4.0

1.4.0 (2025-03-27)

Features Added

• Enhanced binary evaluation results with customizable thresholds

  • Added threshold support for QA and ContentSafety evaluators
  • Evaluation results now include both the score and threshold values
  • Configurable threshold parameter allows custom binary classification boundaries
  • Default thresholds provided for backward compatibility
  • Quality evaluators use "higher is better" scoring (score ≥ threshold is positive)
  • Content safety evaluators use "lower is better" scoring (score ≤ threshold is positive)

• New Built-in evaluator called CodeVulnerabilityEvaluator is added.

  • It provides capabilities to identify the following code vulnerabilities.
    • path-injection
    • sql-injection
    • code-injection
    • stack-trace-exposure
    • incomplete-url-substring-sanitization
    • flask-debug
    • clear-text-logging-sensitive-data
    • incomplete-hostname-regexp
    • server-side-unvalidated-url-redirection
    • weak-cryptographic-algorithm
    • full-ssrf
    • bind-socket-all-network-interfaces
    • client-side-unvalidated-url-redirection
    • likely-bugs
    • reflected-xss
    • clear-text-storage-sensitive-data
    • tarslip
    • hardcoded-credentials
    • insecure-randomness
  • It also supports multiple coding languages such as (Python, Java, C++, C#, Go, Javascript, SQL)

• New Built-in evaluator called UngroundedAttributesEvaluator is added.

  • It evaluates ungrounded inference of human attributes for a given query, response, and context for a single-turn evaluation only,

  • where query represents the user query and response represents the AI system response given the provided context.

  • Ungrounded Attributes checks for whether a response is first, ungrounded, and checks if it contains information about protected class

  • or emotional state of a person.

  • It identifies the following attributes:

    • emotional_state
    • protected_class
    • groundedness

• New Built-in evaluators for Agent Evaluation (Preview)

  • IntentResolutionEvaluator - Evaluates the intent resolution of an agent's response to a user query.
  • ResponseCompletenessEvaluator - Evaluates the response completeness of an agent's response to a user query.
  • TaskAdherenceEvaluator - Evaluates the task adherence of an agent's response to a user query.
  • ToolCallAccuracyEvaluator - Evaluates the accuracy of tool calls made by an agent in response to a user query.

Bugs Fixed

• Fixed error in GroundednessProEvaluator when handling non-numeric values like "n/a" returned from the service.
• Uploading local evaluation results from evaluate with the same run name will no longer result in each online run sharing (and bashing) result files.