---
feat: add etags for cluster and node pool update operations
Clients may now optionally specify the previously read etag for a cluster or node pool during update requests. The etag is regenerated every time the object is updated. If an etag is provided, it will be checked against the current etag, and the request will be rejected on mismatch. This is particularly useful for allowing two clients to concurrently update a repeated field without overwriting each other.
PiperOrigin-RevId: 500787735
Clients may now optionally specify the previously read etag for a cluster or node pool during update requests. The etag is regenerated every time the object is updated. If an etag is provided, it will be checked against the current etag, and the request will be rejected on mismatch. This is particularly useful for allowing two clients to concurrently update a repeated field without overwriting each other.
PiperOrigin-RevId: 498249600
`CLUSTER_SCOPE` option now available in DNSScope
---
feat: add WindowsNodeConfig field to v1alpha1, v1beta1, v1
---
feat: add EphemeralStorageLocalSsdConfig and LocalNvmeSsdBlockConfig APIs to v1alpha1, v1beta1, v1
PiperOrigin-RevId: 497252854
Clients can now configure dual stack clusters by specifying a stack type of IPV4_IPV6 during cluster creation, or with an update command. When used with a cluster creation command that creates a new subnet, the ipv6_access_type field can be used to specify whether the subnet has internal or external IPv6 access. When used with a cluster update command, the subnet must already be dual stack.
PiperOrigin-RevId: 495981345
Clients can now configure dual stack clusters by specifying a stack type of IPV4_IPV6 during cluster creation, or with an update command. When used with a cluster creation command that creates a new subnet, the ipv6_access_type field can be used to specify whether the subnet has internal or external IPv6 access. When used with a cluster update command, the subnet must already be dual stack.
PiperOrigin-RevId: 495438580
chore: disallow "transport=rest" for services where numeric enums are not confirmed to be supported (except in PHP and Java)
PiperOrigin-RevId: 493113566
Use a compact placement policy to specify that nodes within the node pool should be placed in closer physical proximity to each other within a zone. Having nodes closer to each other can reduce network latency between nodes, which can be useful for tightly-coupled batch workloads.
PiperOrigin-RevId: 488490422
---
feat: GKE cluster's control plan/node-pool network isolation
Users now can change the network isolation for your public cluster's control plane.
By default, when a new public cluster is created, GKE assigns a public IP address (external endpoint) to the control plane and provisions
public nodes. New APIs allows to change cluster network isolation of the control plane and/or node-pool from the internet
---
feat: add nodeconfig resource_labels api
---
feat: add API to enable GKE Gateway controller
PiperOrigin-RevId: 486164478
---
feat: GKE cluster's control plan/node-pool network isolation
Users now can change the network isolation for your public cluster's control plane.
By default, when a new public cluster is created, GKE assigns a public IP address (external endpoint) to the control plane and provisions
public nodes. New APIs allows to change cluster network isolation of the control plane and/or node-pool from the internet
---
feat: add nodeconfig resource_labels api
---
feat: add API to enable GKE Gateway controller
PiperOrigin-RevId: 486164143
This RPC is not supposed to be published at this moment because Java generator cannot support this path template.
It will be re-published when the generator is fixed.
---
feat: add APIs for GKE Control Plane Metrics
---
feat: add APIs for GKE node pool blue-green upgrade
---
feat: Add support for output only fields for dual stack clusters
The output only ipv6 fields will be populated for dual stack clusters and it will be seen in "clusters describe" commands.
---
feat: add GetFeatureBlockers API to get the blockers for migrating the workloads to a GKE cluster with the given feature
---
feat: Update Node Pool Update Strategy Enum
---
feat: launch GKE Cost Allocations configuration to the v1beta1 GKE API
---
feat: add a GPU timesharing related API
---
feat: Add support for specifying stack type for clusters
Clients can now configured dual stack clusters by specifying a stack type of IPV4_IPV6. When used with a cluster creation command that creates a new subnet, the ipv6_access_type field can be used to specify whether the subnet has internal or external IPv6 access..
---
feat: add protect config audit api
---
feat: update proto for high throughput logging feature
---
feat: add NCCL fast socket installer API
---
feat: Add Binauthz support to autopilot
---
feat: add etag support to clusters and node pools
---
feat: add visibility label for GPU time-sharing and GPU mps
---
feat: add total size limits to API
---
feat: added support for newly created node-pools to opt in to a premium bandwidth feature, allowing higher bandwidth for VM to VM traffic
---
feat: add support to modify kubelet pod pid limit in node system configuration
---
feat: expand PodAutoscaling message visibility
---
feat: introduce PodAutoscaling message with direct_metrics_opt_in flag
---
feat: add Location Policy API
---
feat: Add Blue/Green Update API
---
feat: add network tags to autopilot cluster
---
feat: Add version EOL to ServerConfig
---
fix: made node pool logging config a node config
---
docs: fix broken link
---
docs: Add busy polling options to LinuxNodeConfig options
---
docs: Fix bulleted list formatting
---
docs: added note about Autopilot not supporting NodeConfig
---
docs: WORKLOADS is deprecated as an allowed value for --monitoring
---
docs: AutoprovisioningNodePoolDefaults.min_cpu_platform field is marked as deprecated
PiperOrigin-RevId: 463701285
This RPC is not supposed to be published at this moment because Java generator cannot support this path template.
It will be re-published when the generator is fixed.
---
feat: add APIs for GKE Control Plane Metrics
---
feat: add APIs for GKE node pool blue-green upgrade
---
feat: add GetFeatureBlockers API to get the blockers for migrating the workloads to a GKE cluster with the given feature
---
feat: Update Node Pool Update Strategy Enum
---
feat: add a GPU timesharing related API
---
feat: add ConfidentialNodes node pool API to v1
---
feat: update proto for high throughput logging feature
---
feat: add V1 GPU sharing config to API
---
feat: add NCCL fast socket installer API
---
feat: Add Binauthz support to autopilot
---
feat: add etag support to clusters and node pools
---
feat: add total size limits to API
---
feat: Add managed prometheus feature to GKE public API
---
feat: added support for newly created node-pools to opt in to a premium bandwidth feature, allowing higher bandwidth for VM to VM traffic
---
feat: add support to modify kubelet pod pid limit in node system configuration
---
feat: add update support for node pool tags, labels and taints
---
feat: add Location Policy API
---
feat: Add Blue/Green Update API
---
feat: add network tags to autopilot cluster
---
feat: Add version EOL to ServerConfig
---
feat: Add CompleteNodePoolUpgradeRequest
---
feat: added spot vm support on GKE
---
feat: GKE Identity Service launches to GA phase
---
fix: made node pool logging config a node config
---
docs: fix broken link
---
docs: Add busy polling options to LinuxNodeConfig options
---
docs: Fix bulleted list formatting
---
docs: added note about Autopilot not supporting NodeConfig
---
docs: AutoprovisioningNodePoolDefaults.min_cpu_platform field is marked as deprecated
PiperOrigin-RevId: 463701283
feat: support spot VM
feat: support Tier 1 bandwidth
feat: update support for node pool labels, taints and network tags
feat: add Binauthz Evaluation mode support to GKE Classic
feat: add GKE Identity Service
feat: add network tags to autopilot cluster
feat: support enabling Confidential Nodes in the node pool
feat: support node pool blue-green upgrade
feat: add Location Policy API
feat: support GPU timesharing
feat: add managed prometheus feature
PiperOrigin-RevId: 458323987
feat: add support to modify kubelet pod pid limit in node system configuration
feat: support Tier 1 bandwidth
feat: support IPV6 Dual Stack (stack_type)
feat: add Binauthz Evaluation mode support to GKE Classic
feat: support GKE Cost Allocations
feat: add network tags to autopilot cluster
feat: add protect config audit api
feat: support enabling Confidential Nodes in the node pool
feat: support node pool blue-green upgrade
feat: add Location Policy API
feat: support GPU timesharing
PiperOrigin-RevId: 458322449
This RPC is not supposed to be published at this moment because Java generator cannot support this path template.
It will be re-published when the generator is fixed.
PiperOrigin-RevId: 423936238
