Google Cloud SDK 220.0.0 introduces several breaking changes, including default disabling of legacy metadata APIs for new Kubernetes Engine node pools and deprecation of the internal gsutil implementation for Python 3 support. Key security enhancements include default secure node pool configurations and KMS encryption for Dataproc clusters. New features span across various services, adding enhanced logging, Python version specification for Composer, and DML support for Cloud Spanner queries.

Jump to: Annotated Release Notes | Stats 📊

Breaking Changes

• Modified output of gcloud container clusters list for DEGRADED clusters to include reason for degradation.
  • File: list.py:1
• Starting in 1.12, new Kubernetes Engine node pools (and default node pools in new clusters) will be created with their legacy Compute Engine instance metadata APIs disabled by default. The flag --metadata disable-legacy-endpoints=true can be used to explicitly create a new node pool (or default pool in a new cluster) with legacy metadata APIs disabled.
  • File: create.py:1
• The gsutil implementation for specific commands is now deprecated to support Python 3. Commands that previously called gsutil internally (e.g., functions deploy, compute images import, dataproc jobs submit variants, composer environments storage commands) now use a different implementation. The storage/use_gsutil property can temporarily restore the old behavior but will eventually be removed.
  • File: gsutil_client.py:1

Security Updates

• Starting in 1.12, new Kubernetes Engine node pools (and default node pools in new clusters) will have their legacy Compute Engine instance metadata APIs disabled by default to enhance metadata protection. Users can explicitly disable them with --metadata disable-legacy-endpoints=true. [HIGH]
  • File: create.py:1
• Added the --gce-pd-kms-key flag to dataproc clusters create to enable protecting clusters with Google Cloud KMS encryption for persistent disks. [MEDIUM]
  • File: create.py:1

New Features by Service

Cloud SDK

• Commands that upload to Google Cloud Storage can now control the upload/download chunksize using the storage/chunk_size property.
  • File: upload.py:1
• Added functionality to gcloud beta help that allows running a search for terms of interest within the help text of gcloud commands.
  • File: gcloud_main.py:1

Cloud Build

• Added a warning message to gcloud builds submit for builds submitted with configs that don’t specify a logging option.
  • File: submit.py:1

Cloud Composer

• Added the --python-version flag to gcloud beta composer environments create to specify the Python version used within the created environment.
  • Flags: --python-version
  • File: create.py:1

Cloud Dataproc

• Added the --gce-pd-kms-key flag to dataproc clusters create to enable protecting clusters with Google Cloud KMS encryption.
  • Flags: --gce-pd-kms-key
  • File: create.py:1

Cloud Key Management Service

• Modified gcloud kms locations list to display information about the availability of Hardware Security Modules in each location.
  • File: list.py:1

Compute Engine

• Added support for managed SSL certificates to gcloud beta compute ssl-certificates.
  • File: ssl_certificates.py:1

Internet of Things

• Added --log-level flag for gcloud iot devices create, gcloud iot devices update, gcloud iot registries create, and gcloud iot registries update for beta.
  • Flags: --log-level
  • File: create.py:1
• Added gcloud iot commands to beta, expanding IoT Core command-line capabilities.
  • File: list.py:1

API Changes

App Engine

• Updated the Java SDK to version 1.9.66.
  • File: java_sdk.py:1
• Updated the Python SDK to version 1.9.77.
  • File: python_sdk.py:1

Cloud Spanner

• Updated gcloud spanner execute-sql to accept DML statements, extending its functionality for database interaction.
  • File: execute_sql.py:1

Stats

• Files changed: 122
• Inserted lines: +6,406
• Deleted lines: -707

Generated by Gemini - View full diff

Google, Google Cloud, and gcloud are trademarks of Google LLC. This analysis was automatically generated by an AI agent examining only public release artifacts, without access to internal or non-public information from Google Cloud.