Archived release notes

Generally available: The accelerator-optimized A3 machine type is now available on Compute Engine. The A3 machine type has NVIDIA® H100 80GB GPUs attached, which can be used to support your large artificial intelligence (AI) models, machine learning (ML), and high performance computing (HPC) workloads. The A3 machine type is available in the following regions and zones:

• APAC
  • Singapore: asia-southeast1-b,c
• Europe
  • Netherlands: europe-west4-b.c
• North America
  • Iowa: us-central1-a,c
  • Virginia: us-east4-a,c
  • Ohio: us-east5-a

Generally available: You can rename an existing VM using the Google Cloud console, gcloud CLI, and REST. For more information, see Rename a VM.

Generally available: NVIDIA L4 GPUs are now available in the following additional regions and zones:

• Seoul, South Korea (asia-northeast3-a)
• Moncks Corner, South Carolina (us-east1-c)

For more information about using GPUs on Compute Engine, see GPU platforms.

Preview: The following quotas and metrics are now available to help you monitor the usage and limits for Compute Engine concurrent operation quotas:

• Quotas for global concurrent operations (metric - compute.googleapis.com/global_concurrent_operations):
  • Concurrent global operations per project
  • Concurrent global operations per project operation type
• Quotas for regional concurrent operations (metric: compute.googleapis.com/regional_concurrent_operations):
  • Concurrent regional operations per project
  • Concurrent regional operations per project operation type

For more information, see Concurrent operation quotas.

Preview: Managed instance groups (MIGs) let you create pools of suspended and stopped virtual machine (VM) instances. You can manually suspend and stop VMs in a MIG to save on costs, or use suspended and stopped pools to speed up scale out operations of your MIG.

For more information, see Work with suspended and stopped VMs in a MIG.

Generally available: The following location and scale enhancements for Persistent Disk Asynchronous Replication are generally available:

• Larger disk capacity: the maximum disk size has increased from 2 TiB to 5 TiB.
• Faster replication rate: the disk data replication rate has increased from 100 MB/min to 250 MB/min.
• Expanded regional support: PD Async Replication is available in 15 additional regions across Europe, APAC, and North America. For the full list of available regions, see Supported region pairs.

Preview: When creating or modifying an on-demand reservation, you can configure reservations to be automatically deleted at a specific date and time. Automatically deleting reservations makes it easier to prevent charges from unused reservations when you no longer need them.

For more information, see the documentation for creating on-demand reservations.

Preview: In a managed instance group (MIG), you can turn off repairs to inspect failed and unhealthy VMs, to implement your own repair logic, or to monitor the application health without triggering repairs by MIG. For more information, see Turn off repairs in a MIG.

Generally available: A replica recovery checkpoint of a regional Persistent Disk volume represents the most recent crash-consistent point in time of the fully replicated disk. For disks that are not fully replicated, you can use the checkpoint to create disk snapshots from an incomplete zonal replica. You can create and use these snapshots to recover disk data in the rare scenario where your synced replica goes down before your incomplete replica catches up.

Learn more about Regional Persistent Disk replica recovery checkpoints and how to use checkpoints to recover a degraded disk.

The h3-node-88-352 sole-tenant node type is now Generally Available.

Generally available: When assigning a custom queue count for the receive and transmit queues for a vNIC, under certain conditions, you can configure a number of custom queue counts that exceeds the number of vCPUs allocated to the VM.

Preview: Advanced maintenance control for sole-tenancy lets you control planned maintenance events for sole-tenant node groups and minimise maintenance-related disruptions. This feature is available only for sole-tenant node groups. To use this feature with your existing virtual machines, you must first move your VMs to sole-tenant node groups that have advanced maintenance control enabled.

The advanced maintenance control for sole-tenancy feature lets you:

• Check for maintenance events scheduled for a sole-tenant node 28 days in advance.
• Trigger maintenance immediately or schedule it for later. Note that if you trigger maintenance immediately, the maintenance takes place within 24-hours from the time you trigger the request.

For more information, see Advanced maintenance control for sole-tenancy.

Preview: Hyperdisk Balanced is now available in preview with H3 VMs. Hyperdisk Balanced is a good fit for a wide range of use cases such as LOB applications, web applications, and medium-tier databases that don't require the performance of Hyperdisk Extreme. For more information, see About Hyperdisk.

Preview: Project zonal metadata is custom project metadata that you can set exclusively for VMs in a specific zone in a project. Project zonal metadata helps you with fault isolation and provides greater reliability. By setting custom project zonal metadata, you gain more control over the project metadata for your VMs and limit the impact of any incorrect metadata updates to VMs within the specific zone.

Learn more about VM metadata and how to set custom project zonal metadata.

Generally available: c3d-standard, c3d-highmem, c3d-highcpu, and c3d-standard-lssd machine types for general-purpose C3D VMs are generally available.

Generally available: C3 VMs support Compute Engine flexible committed use discounts (CUDs).

Compute Engine flexible CUDs allow you to commit to a minimum hourly spend amount and use vCPUs and/or memory in any of the projects within your Cloud Billing account, across any region, and belonging to any eligible machine types. Learn more about Compute Engine Flexible CUDs and how to purchase flexible commitments.

If you want to modify a future reservation request using the Compute Engine API, the paths query parameter is deprecated. Instead, use the updateMask query parameter.

For more information, see Modify future reservation requests.

Preview: You can now use workforce identity federation with OS Login.

Preview: The following metrics are now available to help you monitor your Persistent Disk and Hyperdisk volume performance:

• Average I/O latency (compute.googleapis.com/instance/disk/average_io_latency)

• Average I/O queue depth (compute.googleapis.com/instance/disk/average_io_queue_depth)

To learn more about these metrics and how to view them, see Review disk metrics.

Generally available: You can configure stateful IP addresses in a managed instance group. Stateful IP addresses are preserved when VM instances in the group are repaired, updated, and re-created. For more information, see Configuring stateful IP addresses in MIGs.

Generally available: H3 VMs, designed for compute-intensive high performance computing (HPC) workloads, are now generally available. For more information, see H3 machine series.

When you install the Ops Agent on a Compute Engine VM by using the Observability tab on a Compute Engine VM details page, the agent is now installed with an Ops Agent OS policy. This installation method replaces the prior set of manual steps. For more information, see Installing the agent by using the Google Cloud console.

Generally available: NVIDIA L4 GPUs are now available in the following additional regions and zones:

• APAC
  • Seoul, South Korea (asia-northeast3-b)
• Europe
  • St. Ghislain, Belgium (europe-west1-b)
  • Frankfurt, Germany (europe-west3-b)
• North America
  • Council Bluffs, Iowa: (us-central1-c)
  • Las Vegas, Nevada (us-west4-a,c)

For more information about using GPUs on Compute Engine, see GPU platforms.

Generally available: NVIDIA L4 GPUs are now available in the following additional regions and zones:

• Singapore(asia-southeast1-a)

For more information about using GPUs on Compute Engine, see GPU platforms.

Creating a reservation or future reservation request by using an instance template that specifies an A2, C3, or G2 machine type causes errors or problems with consumption. For more information, see Known issues.

Preview: c3d-standard, c3d-highmem, c3d-highcpu, and c3d-standard-lssd virtual machines are available in the following regions:

• Council Bluffs, Iowa, North America, us-central1
• Moncks Corner, South Carolina, North America, us-east1
• Ashburn, Virginia, North America , us-east4
• St. Ghislain, Belgium, Europe, europe-west1
• Eemshaven, Netherlands, Europe, europe-west4
• Jurong West, Singapore, Asia, asia-southeast1

See the General purpose machines document for details.

Preview: You can now view the organization-wide patch status dashboard and OS policy compliance reports by using VM Manager.

Preview: Compute Engine API now enforces the Filtered list cost overhead quota, which limits the number of resources to be filtered out from server-side *.list and *.aggregatedList methods.

The quota is charged against the following metrics:

• Global: compute.googleapis.com/filtered_list_cost_overhead
• Regional: compute.googleapis.com/filtered_list_cost_overhead_per_region

For more information, see Rate quotas and best practices for list filtering.

The Google Cloud console labels for OS patch management and OS configuration management on VM Manager pages have been renamed to Patch and OS policies respectively.

Generally available: Autohealing in managed instance groups (MIG) supports regional health checks. To reduce cross-region dependency or to achieve data residency, use a regional health check. For more information, see Set up an application health check and autohealing.

Generally available: Instance templates are available as both regional and global resources. To reduce cross-region dependency or to achieve data residency, use a regional instance template to create virtual machines (VM), managed instance groups (MIG), or reservations. For more information, see Regional and global instance templates.

Generally available: Dammam, Kingdom of Saudi Arabia, Middle East me-central2-a,b,c has launched with E2, N2, N2D, and T2D VMs in all three zones. See the Dammam region access document to learn more.

Preview: Snapshot settings are centralized configuration parameters for all snapshots in a project. You can use snapshot settings to customize the default storage location for all future snapshots in your project. By enabling you to do this, snapshot settings remove the need for you to manually specify a storage location during each individual snapshot creation.

Learn more about snapshot settings and how to set the default storage location for a project using snapshot settings.

Generally available: The Red Hat Knowledgebase provides you with access to articles, solutions, product documentation, and community discussions for Red Hat products.

You can now access the Red Hat Knowledgebase by using single-sign-on (SSO) through the Google Cloud console from your Red Hat Enterprise Linux (RHEL) VMs. For more information, see Access Red Hat Knowledgebase.

Generally available: You can create C3-standard VMs with Local SSD attached using new machine types, for example c3-standard-44-lssd. For more information, see Choosing a valid number of Local SSDs.

You can manage future reservations using the Google Cloud console. Future reservations provide a high level of assurance to obtain important or difficult-to-obtain capacity in advance.

For more information, see the following pages:

• Create future reservation requests for a single project
• Create shared future reservation requests
• Delete future reservation requests
• Modify future reservation requests
• View future reservation requests

Generally available: NVIDIA L4 GPUs are now available in the following additional regions and zones:

• APAC
  • Taiwan (asia-east1-a,c)
  • Tokyo (asia-northeast1-a,c)
  • Singapore(asia-southeast1-c)
• Europe
  • Belgium (europe-west1-c)
  • London (europe-west2-a,b)
• North America
  • Northern virginia (us-east4-c)
  • Dalles (us-west1-c)

For more information about using GPUs on Compute Engine, see GPU platforms.

Preview: You can use future reservations to reserve resources at a specific date and time in the near future.

Future reservations are useful for obtaining capacity for future peak-demand events or highly-demanded resources. For more information, see About future reservation requests.

Preview: You can reduce network latency between VMs by using compact placement policies to specify the maximum distance between VMs. Use compact placement policies to optimize workloads with frequent communication across VMs—for example, high-performance computing (HPC), machine learning (ML), or database server workloads. You can keep VMs in the same rack, across adjacent racks within the same cluster, or across adjacent clusters.

For more information, see Reduce latency by using compact placement policies.

Generally available: Berlin, Germany, Europe europe-west10-a,b,c has launched with E2, N2, N2D, and T2D VMs available in all three zones.

Generally available: When a managed instance group (MIG) repairs a failed or an unhealthy VM, you can apply the latest instance template and per-instance configuration to recreate the VM instead of applying the configuration originally used to create the VM. For more information, see Apply configuration updates during repairs.

Generally available: Hyperdisk Throughput is now available in the following additional regions and zones:

• Council Bluffs, Iowa: us-central1
• Moncks Corner, South Carolina: us-east1
• Ashburn, Virginia: us-east4-b, c
• Eemshaven, Netherlands: europe-west4-a, c
• Jurong West, Singapore: asia-southeast1
• Mumbai, India: asia-south1-a

Generally available: The Ops Agent (version 2.38.0 and later) now supports the automatic tracking of GPU usage metrics reported from the NVIDIA Management Library (NVML) for Linux virtual machine instances that have attached NVIDIA GPUs.

Through an available integration with NVIDIA's Data Center GPU Manager (DCGM), you can also track metrics such as Streaming Multiprocessor (SM) block utilization, SM occupancy, SM pipe utilization, PCIe traffic rate, and NVLink traffic rate.

For more information, see Monitoring GPU performance on Linux VMs.

Generally available: You can use Cloud Monitoring to monitor the consumption of your reservations and set custom alerts. For more information, see Monitor consumption of reservations.

Generally available: If a host error occurs on a VM, you can control how much time Compute Engine spends recovering Local SSD data with the Local SSD recovery timeout setting. For more information, see Local SSD data persistence.

Generally available: Use the new distribution shape ANY SINGLE ZONE in a regional Managed Instance Group (MIG) to automatically select a single zone that has available resources within your quota. Recommended for workloads that require low latency, high-bandwidth connections between VMs or when you want to avoid inter-zone network traffic costs.

Preview: You can create C3 VMs with Local SSD attached using new machine types (-lssd). For more information, see Add a local SSD to your VM.

Preview: H3 VMs, designed for compute-intensive high performance computing (HPC) workloads, are now in preview. For more information, see H3 machine series.

Generally available: c3-standard and c3-highmem machine types for general-purpose C3 VMs are generally available.

Generally available: You can modify the description, schedule frequency, retention policy, or labels for a snapshot schedule instead of creating a new snapshot schedule. For more information, see Change a snapshot schedule.

Preview: You can now use SSH-in-browser to connect to TPU VMs.

Generally available: NVIDIA T4 GPUs are now available in the following additional regions and zones:

• Salt Lake City: us-west3-b

For more information about using GPUs on Compute Engine, see GPU platforms.

Generally available: You can enable faster network packet processing by using the Data Plane Development Kit (DPDK). DPDK helps you to optimize VMs that run network-intensive workloads, such as video streaming or voice calls.

Preview: You can use instant snapshots to take in-place disk backups that can be restored to new disks under a minute.

Instant snapshots are ideal for rapid data restoration within the same location as the source disk. For more information, see Instant snapshots.

Generally available: You can now use a regional Persistent Disk as a VM boot disk.

You can suspend and resume E2 VMs.

Preview: c3-standard and c3-highmem machine types are now available for general-purpose C3 VMs.

Generally Available: Persistent Disk Asynchronous Replication (PD Async Replication) is now generally available. For more information, see About Persistent Disk Asynchronous Replication.

Generally available: NVIDIA A100 80GB GPUs are now available in the following additional regions and zones:

• Ohio, North America: us-east5-b

For more information about using GPUs on Compute Engine, see GPU platforms.

Generally available: For managed instance groups (MIGs), Google Cloud Console provides you with an improved way to configure autoscaling based on Cloud Monitoring metrics. The redesigned user interface enables you to explore available metrics and filters. You can visualize the metric values in a chart, which also displays the aggregated value used for autoscaling.

Preview: You can now use custom constraints to provide more granular and customizable control over specific fields for some Compute resources. For more information, see Manage Compute Engine resources using custom constraints.

Generally available: Hyperdisk Throughput provides cost-effective and throughput-oriented block storage with dynamically configurable capacity and throughput. Hyperdisk volumes are durable network storage devices that your VMs can access, similar to Persistent Disk. For more information, see About Hyperdisk.

Generally available: NVIDIA A100 80GB GPUs are now available in the following additional regions and zones:

• Iowa, North America: us-central1-a

For more information about using GPUs on Compute Engine, see GPU platforms.

For MIGs that have T2D machine series VMs, autoscaling based on CPU utilization doesn't work as expected. For more details, see Known issues.

Generally available: Accelerator-optimized (G2) machine types with attached NVIDIA® L4 GPUs are generally available in the following regions and zones:

• Singapore, APAC: asia-southeast1-b
• Netherlands, Europe: europe-west4-a,b,c
• Iowa, North America: us-central1-a,b
• South Carolina, North America: us-east1-b,d
• Virginia, North America: us-east4-a
• Oregon, North America: us-west1-a,b

Preview: In a managed instance group (MIG), you can set metadata and labels for all VMs in the group without the need to create a new instance template. For more information, see Override instance template properties with an all-instances configuration.

The image import tool now supports importing CentOS Stream 9 and CentOS Stream 8 images to Google Cloud.

Generally available: NVIDIA A100 80GB GPUs are now available in the following additional regions and zones:

• Netherlands, Europe: europe-west4-a
• Singapore, APAC: asia-southeast1-c

For more information about using GPUs on Compute Engine, see GPU platforms.

Generally available: General purpose C3 VMs are now generally available in the following regions:

• Council Bluffs, Iowa, North America : us-central1
• Moncks Corner, South Carolina, North America: us-east1
• Ashburn, Virginia, North America: us-east4
• St. Ghislain, Belgium, Europe: europe-west1
• Eemshaven, Netherlands, Europe : europe-west4
• Jurong West, Singapore, APAC: asia-southeast1

Preview: You can now use the discard-local-ssd=false flag to preserve the contents of a single attached Local SSD disk when suspending or stopping a VM. For more information, see the Local SSD Documentation.

The image import tool now supports importing Rocky Linux 9 images to Google Cloud.

Generally available: The local SSD quota per machine family (LOCAL_SSD_TOTAL_GB_PER_VM_FAMILY) is generally available. Use the quota metric compute.googleapis.com/local_ssd_total_storage_per_vm_family instead of compute.googleapis.com/local_ssd_total_storage to view the quota usage and limits for local SSD in your project. For more information, see View and manage local SSD quota per machine family.

End of life: On May 1, 2024, NVIDIA K80s will be end of life and won't be available for new or existing VMs on Google Cloud.

For information about how to prepare for this EOL, see NVIDIA K80 EOL.

Two vulnerabilities (CVE-2023-1017 and CVE-2023-1018) were discovered in Trusted Platform Module (TPM) 2.0.

For more information, see the GCP-2023-004 security bulletin.

In the Google Cloud console, the Observability tab on the VM instances page for Compute Engine has been enhanced. Disk and Network sections with additional charts have been added. The Integrations > Detected section lets you navigate to the dashboards for the third-party integrations that you have configured, like Apache or NGINX. The page also includes a set of recommended alerts for setting up pre-configured alerting policies for CPU, memory, and disk utilization and for host errors.

You can now create regional Persistent Disk volumes when creating a new VM either directly, or through instance templates. For more information, see Create a VM instance with additional non-boot disks or Create a new instance template.

Preview:

• The HPC Rocky Linux 8 image is now available for HPC workloads.
• The HPC VM Images now support Intel MPI 2021 with tools to easily installing the Intel MPI 2021 library, the net and psm3 libfabric providers.
• The HPC VM Images now support OpenMPI. For more details, see Open MPI best practice guides.

Generally available: You can now use the gcloud command-line tool to import images from AWS into Google Cloud. For more information, see Importing images from AWS.

Preview: Accelerator-optimized (G2) machine types are now available on Compute Engine. Each G2 machine type has a fixed number of NVIDIA® L4 GPUs attached to support your next generation graphics performance workloads. The G2 machine types are available in the following three regions:

• Iowa, North America: us-central1-a,b
• Netherlands, Europe: europe-west4-a
• Singapore, APAC: asia-southeast1-b

Generally available: You can use the Regional disk replica state metric in Cloud Monitoring to track the states of your regional Persistent Disk zonal replicas. You can also use the metric data to determine the replication state of your regional Persistent Disk volumes.

Learn more about zonal replication for regional Persistent Disk and how to monitor the states of regional Persistent Disk zonal replicas.

Preview: Persistent Disk Asynchronous Replication (PD Async Replication) provides low recovery point objective (RPO) and low recovery time objective (RTO) block storage replication for cross-region active-passive disaster recovery. For more information, see About Persistent Disk Asynchronous Replication.

Generally available: Doha, Qatar, Middle East me-central1-a,b,c has launched with E2 and N2 VMs available in all three zones.

See VM instance pricing for details.

Generally Available: You can test how workloads running on sole-tenant nodes behave during a host maintenance event, and see the effects of the sole-tenant VM's host maintenance policy on the applications running on the VMs.

For more information, see Simulate host maintenance events on sole-tenant nodes.

Generally available: Turin, Italy, Europe europe-west12-a,b,c has launched with E2, N2, N2D, and T2D VMs available in all three zones. See VM instance pricing for details.

Your automated processes might fail if they use API response data about your resource-based commitment quotas. For more information, see Known issues.

End of life: On May 31, 2023, Ubuntu 18.04 LTS (Bionic) will reach end of life and the images deprecated on Google Cloud. If you use Ubuntu 18.04 LTS images in your project, review Ubuntu LTS end of life.

Generally available: Hyperdisk provides the fastest block storage for Compute Engine for your high-end, memory intensive workloads. Hyperdisk volumes are durable network storage devices that your VMs can access, similar to Persistent Disk. For more information, see About Hyperdisk.

Generally available: When creating a reservation, you can now include a compact placement policy to specify that VMs should be located as close to each other as possible to reduce network latency. Learn how to create a reservation that specifies a compact placement policy.

Generally available: You can upgrade the term of your 1-year commitments and convert them into 3-year commitments to get a higher discount percentage for your committed resources and continue receiving the discounts for a longer time period.

For more information, see Upgrade the term of commitments.

Generally available: The image import tool now supports importing SUSE Linux Enterprise Server 15 SP4 and SUSE Linux Enterprise Server 15 SP4 for SAP images to Google Cloud.

Generally available: NVIDIA® T4 GPUs are now available in the following region and zones:

• Warsaw, Poland, Europe: europe-central2-b,c

For more information about using GPUs on Compute Engine, see GPU platforms.

Regional metrics for Compute Engine API limits are now available. Regional migration of API limits reduces the scope of global or multi-regional outages. For more information about the new regional metrics and changes in API limits, see API rate limits.

Due to this change, you might want to update your Cloud Monitoring dashboards, queries and alerts to use the regional metrics. For more information, see Migrate Compute Engine API quota from global metrics to regional metrics.

Preview: You can autoscale a regional managed instance group with a BALANCED target distribution shape. With the BALANCED shape, the autoscaler is aware of the capacity in each zone and creates VMs in zones that have resource availability. For more information, see Autoscaling a regional MIG.

Preview: C3 VMs are now available in the following regions:

• Council Bluffs, Iowa, North America : us-central1
• Ashburn, Virginia, North America: us-east4
• Eemshaven, Netherlands, Europe : europe-west4

Preview: You can now use a GPU-enabled Ops Agent to track GPU utilization and GPU memory usage rates for Linux virtual machine instances that have attached GPUs.

Through an available integration with NVIDIA's Data Center GPU Manager (DCGM), you can also track metrics such as Streaming Multiprocessor (SM) block utilization, SM occupancy, SM pipe utilization, PCIe traffic rate, and NVLink traffic rate.

For more information, see Monitoring GPU performance on Linux VMs.

Tau T2A VMs now support secure boot.

Preview: You can modify the description, schedule frequency, or labels for a snapshot schedule instead of creating a new snapshot schedule. For more information, see Change a snapshot schedule.

Generally available: You can now use an instance template to define the properties of a reservation and the VMs that can consume the reservation in the same place. Learn how to create a reservation by specifying an instance template.

Generally available: Compute Engine committed use discounts are available for Red Hat Enterprise Linux (RHEL) image licenses. Learn more about discounted RHEL image pricing and how to purchase a license commitment.

You can now use the Observability tab on the Compute Engine VM instances page to see the five virtual machines consuming the most of a resource. For more information, see Troubleshooting VM performance issues.

Preview: When a managed instance group (MIG) repairs a failed or an unhealthy VM, you can apply the latest instance template and per-instance configuration to recreate the VM instead of applying the configuration originally used to create the VM. For more information, see Apply configuration updates during repairs.

Preview: You can now simulate host maintenance events on sole-tenant nodes.

For more information, see Simulate host maintenance events on sole-tenant nodes.

Preview: Use the Google Cloud console to rename VMs. For more information, see Rename a VM.

Generally available: N2 VMs with 64 or more vCPUs now support up to 4 GB/s (read) and 3 GB/s (write) throughput per instance with Extreme persistent disks (pd-extreme). Previously the maximum was 2.2 GB/s per instance.

The image import tool now supports importing RHEL 9 images to Google Cloud.

Generally available: NVIDIA® T4 GPUs are now available in the following region and zones:

• Hong Kong, APAC: asia-east2-a,c

For more information about using GPUs on Compute Engine, see GPU platforms.

Generally available: You can merge your active hardware resource commitments into one larger commitment to track and manage them as a single entity. You can now also merge your commitments by using the Google Cloud Console. For more information, see Merging commitments.

Preview: You can limit the runtime of a VM to automatically stop or delete it when a time limit is reached. Limiting VM runtimes can help you optimize temporary workloads by minimizing costs and releasing quota. For more information, see Limit the runtime of a VM.

Generally available: You can double the default size limit for a managed instance group (MIG): Zonal MIGs support up to 2,000 VMs and regional MIGs support up to 4,000 VMs. For more information, see Increase the group's size limit.

Preview: Use the new distribution shape ANY SINGLE ZONE in a regional managed instance group (MIG) to automatically select a single zone that has available resources within your quota. Recommended for workloads that require low latency, high-bandwidth connections between VMs or when you want to avoid inter-zone network traffic costs.

Balanced persistent disks and SSD persistent disks now offer baseline IOPS and throughput performance. To learn more, see Baseline performance.

Generally available: Share sole-tenant node groups with other projects or with your entire organization. For more information, see Share sole-tenant node groups.

Per VM Tier_1 networking performance now includes up to 25 Gbps egress for traffic going to public IP addresses (increased from 7 Gbps).

The quota limits displayed in the Cloud console might be incorrect in the asia-south1 region. For more information, see Known issues.

Generally available: Memory-optimized M3 virtual machine instances are available in the following regions and zones:

• Frankfurt, Germany (europe-west3-a,b)
• Eemshaven, Netherlands (europe-west4-a,b)
• Council Bluffs, Iowa, USA (us-central1-a,b)
• Las Vegas, Nevada, USA (us-west4-a,b)

See VM instance pricing for details.

The image import tool now supports importing Ubuntu 22.04 LTS and Windows 11 images to Google Cloud.

Generally available: Compute Engine flexible committed use discounts (flexible CUDs) are spend-based discounts that add flexibility to your spending capabilities by eliminating the need to restrict your commitments to a single project, region, or machine series. You can purchase flexible commitments and commit to a minimum hourly spend amount to use vCPUs and/or memory in any of the projects within your Cloud Billing account, across any region, and belonging to any eligible general-purpose and/or compute-optimized machine types.

Learn more about flexible CUDs and how to purchase flexible commitments.

Generally available: You can resize an existing hardware resource commitment and split it into smaller commitments to closely monitor and manage portions of one large commitment in the form of smaller individual commitments. You can now also split your commitments by using the Google Cloud Console. For more information, see Splitting commitments.

Generally available: Accelerator-optimized (A2 ultraGPU) machine types with their attached A100 80GB GPUs are now available in the following regions and zones:

• Iowa, North America: us-central1-c
• Ashburn, Virginia, North America: us-east4-c

The incorrect quota limits displayed in the Cloud console in the me-west1 region have been resolved.

Generally available: Tau T2A, Google Cloud's first general purpose VM family to run on Arm architecture, is now generally available in these three regions.

For more information, including how to try T2A for free for a limited time, see Creating an Arm VM instance.

Generally available: View the VM placement topology information to determine how close a VM is located in relation to another VM. For more information, see View VM placement topology.

Generally available: Reduce licensing costs by customizing the number of visible CPU cores.

Generally Available: E2 shared-core custom VMs are now generally available. See VM instance pricing for details.

The quota limits displayed in the Cloud console might be incorrect in the me-west1 region. For more information, see Known issues.

Generally available: A new machine type for the memory-optimized-machine family called m2-hypermem-416 with 416 vCPUs and 8832 GB of memory. This new machine type is now generally available in the same regions as the other M2 machine types.

For more information, see Memory-optimized-machine family.

Generally available: NVIDIA® T4 GPUs are now available in the following region and zones in Middle East:

• Tel Aviv, Israel: me-west1-b,c.

For more information about using GPUs on Compute Engine, see GPU platforms.

Generally available: Tel Aviv, Israel, Middle East me-west1-a,b,c has launched with E2 and N2 VMs available in all three zones, and M1 VMs in zones a and c.

See VM instance pricing for details.

Generally available: Compute Engine supports importing a virtual disk with an UEFI bootloader. Learn more about using the --guest-os-features flag to enable UEFI booting for the imported disk.

The incorrect quota limits displayed in the Cloud console in the us-east5 region have been resolved.

Generally available: Archive snapshots are now available for more cost-efficient data retention as compared to regular snapshots, which are best suited for long-term back up and disaster recovery. For more information, see Archive snapshots.

Generally available: To reduce image licensing cost, you can now bring your Red Hat Enterprise Linux subscriptions to Google Cloud. For more information, see Create a VM using a RHEL BYOS image.

Preview: Accelerator-optimized (A2 ultraGPU) machine types with their attached A100 80GB GPUs are now available in the following region:

• Iowa, North America: us-central1-c

The following changes have been introduced to how your resource usage is calculated to determine applicable sustained use discounts:

• Usage will be calculated on an hourly basis instead of a per microsecond basis.
• Usage will be calculated collectively for a billing account instead of on a per project basis.

Preview: You can double the default size limit for a managed instance group (MIG): Zonal MIGs now support up to 2,000 VMs and regional MIGs support up to 4,000 VMs. For more information, see Increase the group's size limit

Generally Available: Internal and external IPv6 addresses for Google Compute Engine instances are available in all regions.

For more information, see Configuring IPv6 for instances and Creating instances with multiple network interfaces.

Generally available: You can now use the os-config troubleshoot command to help verify the setup of VM Manager. For more information, see Verifying VM Manager setup.

Generally available: NVIDIA® T4 GPUs are now available in the following additional regions and zones:

• Ashburn, Virginia, North America: us-east4-a

For more information about using GPUs on Compute Engine, see GPU platforms.

The quota limits displayed in the Cloud console might be incorrect in the us-east5 region. For more information, see Known issues.

Generally available: When you autoscale a MIG, you can view the reasons for why the autoscaler adds or removes VMs in your MIG. For more information, see Viewing autoscaler logs.

Preview: You can now merge or split your existing hardware resource commitments to create new upsized or downsized commitments. For more information, see Merge and split commitments.

Generally available: Use the Cloud console, the gcloud tool, or the API to configure a VM to shut down when a Cloud KMS key is revoked. For more information, see Configure VM shutdown on Cloud KMS key revocation.

Generally available: When you create VMs in bulk, you can now use the following new values with the TARGET_SHAPE flag:

• ANY: Use this value to place VMs in zones to maximize unused zonal reservations.
• BALANCED: Use this value to place VMs uniformly across zones.

Generally available: Compute Engine committed use discounts are now Generally Available for SUSE Linux Enterprise Server (SLES) image licenses. Learn more about discounted SLES image pricing and how to purchase a license commitment.

Generally available: NVIDIA® T4 GPUs are now available in the following additional regions and zones:

• Montréal, Québec, North America : northamerica-northeast1-c

For more information about using GPUs on Compute Engine, see GPU platforms.

Generally available: Internal and external IPv6 addresses for Google Compute Engine instances are available in all regions.

For more information, see Configuring IPv6 for instances and instance templates and Creating instances with multiple network interfaces.

Generally available: You can use the Cloud console to configure autoscaling based on unacknowledged messages in a Pub/Sub subscription. For more information, see Autoscale based on unacknowledged messages in Pub/Sub.

Generally available: NVIDIA® T4 GPUs are now available in the following additional regions and zones:

Ashburn, Virginia, North America : us-east4-c

For more information about using GPUs on Compute Engine, see GPU platforms.

Preview: Tau T2A, Google Cloud's first general purpose VM family to run on Arm architecture, is now available. Tau T2A VMs are available in three regions.

For more information, see Arm VMs on Compute Engine.

Generally Available: A version of Rocky Linux is now available that is optimized for running on Compute Engine.

This version of Rocky Linux is configured to use the latest version of the Google virtual network interface (gVNIC) which is specifically designed to support workloads that require higher network bandwidths. For more information, see the Rocky Linux section of the Operating systems details documentation.

Generally available: You can now create shared reservations of Compute Engine zonal resources using the Google Cloud Console. Learn about shared reservations and creating a shared reservation.

GA: You can now use the SSH troubleshooting tool from the Cloud console to help you determine the cause of failed SSH connections. For more information, see SSH troubleshooting tool.

The CPU utilization observability metric is incorrect for VMs that use one thread per core. For more information, see Known issues.

Preview: Windows VMs now support SSH connections from the gcloud CLI. For more information, see Connect to Windows VMs using SSH.

Cloud console SSH-in-browser connections might fail if you use custom firewall rules. For workarounds, see Known issues.

Generally available: Optimize the distribution of VMs in sole-tenant node groups. For more information, see About manual live migration.

Generally Available: The image import tool now supports importing Windows Server 2022 images to Google Cloud.

Generally Available: Compute Engine can now use a maximum network packet size of 8896 when communicating between VMs on the same subnet. For details, see the maximum transmission unit overview.

Generally available: Dallas, Texas us-south1-a,b,c has launched with E2 and N2 VMs available in all three zones.

See VM instance pricing for details.

Generally available: NVIDIA A100 GPUs are now available in the following additional regions and zones:

Las Vegas, Nevada, North America : us-west4-b

For more information about using GPUs on Compute Engine, see GPU platforms.

Preview: When you create VMs in bulk, you can now use the following new values with the TARGET_SHAPE flag:

• ANY: Use this value to place VMs in zones to maximize unused zonal reservations.
• BALANCED: Use this value to place VMs uniformly across zones.

Generally available: NVIDIA A100 GPUs are now available in the following additional regions and zones:

Seoul, South Korea, APAC : asia-northeast3-a,b

For more information about using GPUs on Compute Engine, see GPU platforms.

Preview: You can now use the SSH troubleshooting tool from the Cloud console to help you determine the cause of failed SSH connections.

Generally available: Columbus, Ohio, USA us-east5-a,b,c has launched with E2, N2, and N2D VMs in all three zones. Additionally, you can create C2 VMs in zones a and b.

See VM instance pricing for details.

N2D VMs are now available in Paris, France europe-west9-a,b,c.

See VM instance pricing for details.

Generally available: You can access Google APIs and services from Compute Engine instances using either internal IPv6 addresses with Private Google access or external IPv6 addresses.

Generally available: Madrid, Spain europe-southwest1-a,b,c has launched with E2 and N2 VMs available in all three zones.

See VM instance pricing for details.

Generally available: Insights for idle VM and machine size recommendations help you assess the utilization of your Compute Engine resources. Insights are automatically generated based on system metrics or metrics gathered by the Cloud Monitoring service.

Learn more about VM insights and MIG insights.

Generally available: Paris, France europe-west9-a,b,c has launched with general-purpose E2 and N2 VMs available in all three zones.

See VM instance pricing for details.

Generally available: Spot VMs are available for all machine types, regions, and zones. Use Spot VMs for workloads that can withstand preemption to receive large discounts. Spot VMs provide discounts of 60-91% off the on-demand price for standard VMs for machine types and GPUs and also provide smaller discounts for local SSDs. Spot prices can change up to once a month to reflect the underlying supply and demand.

Spot VMs are the latest version of preemptible VM instances. Although new and existing preemptible VMs continue to be supported and use the same prices as Spot VMs, Spot VMs provide new features that are not supported for preemptible VMs. For example, preemptible VMs can only run for up to 24 hours at a time, but Spot VMs have no maximum runtime.

Learn more about Spot VMs and preemptible VMs.

NVIDIA 510 driver is now supported for GPUs running on Compute Engine. For information about installing drivers, see Install GPU drivers.

Generally available: Milan, Italy europe-west8-a,b,c region has launched with general-purpose E2, N2, and N2D VMs available in all three zones.

See VM instance pricing for details.

Preview: You can now customize the number of visible CPU cores.

Generally available: NVIDIA A100 GPUs are now available in the following additional regions and zones:

• Tokyo, Japan, APAC: asia-northeast1-a,c

For more information about using GPUs on Compute Engine, see GPU platforms. For pricing information, review the pricing tables for the Accelerator-optimized machine type family.

Tau T2D VMs are now available in the following regions and zones:

• Las Vegas, NV (us-west4-a,b)
• São Paulo, Brazil, South America (southamerica-east1-a,b,c)
• St. Ghislain, Belgium (europe-west1-c)

N2 general-purpose VMs are available in Salt Lake City, UT (us-west3-a,b,c).

See VM instance pricing for details.

Generally available: You can now set the number of threads per core on a VM.

Generally available: You can now autoscale both regional and zonal managed instance groups based on a Cloud Monitoring metric that provides an aggregated value for the group. You can also apply filters to group metrics to further scope the scaling signal. For more information, see Scaling based on Cloud Monitoring metrics.

General purpose Tau T2D VMs have limited availability in London (europe-west2-a,c). If you are interested in trying out T2D, speak to your Google Cloud Account Team. For pricing details, see VM instance pricing.

Preview: You can now share sole-tenant node groups with other projects or your entire organization.

General-purpose Tau T2D virtual machine instances are available in the following regions and zones:

• Northern Virginia (us-east4-a,b,c)
• South Carolina (us-east1-b,c,d)
• Frankfurt (europe-west3-a,b,c)
• Sydney (australia-southeast1-a,b,c)
• Taiwan (asia-east1-a,b,c)

See VM instance pricing for details.

Generally available: Compute Engine now supports Suspend and Resume in General Availability.

Fixed the issue causing the Compute Engine API Quotas page in the Cloud Console to display duplicate API quota groups.

Generally available: NVIDIA® T4 GPUs are now available in the following additional regions and zones:

• Council Bluffs, Iowa, North America : us-central1-c
• Eemshaven, Netherlands, Europe : europe-west4-a

For more information about using GPUs on Compute Engine, see GPUs on Compute Engine.

Public Preview: You can set the maximum amount of time that Compute Engine waits before terminating or restarting an unresponsive VM. For more information, see Set VM availability policies.

NVIDIA 510 driver not yet supported for GPUs running on Compute Engine, see Known issues.

New documentation for licenses and appending licenses.

T2D machines are now available in the following regions and zones:

• St. Ghislain, Belgium: europe-west1
• The Dalles, Oregon: us-west1

See VM instance pricing for details.

Generally available: Compute-optimized C2D machine types are now generally available. C2D machine types are built on top of third generation AMD EPYC Milan processors and are a great fit for high-performance computing (HPC) workloads. For more information, see Compute-optimized machine family.

Public Preview: You can now use the security keys registered for 2-Step Verification in your Google account to connect to VMs that use OS Login. For more information, see Enable security keys with OS Login.

Generally available: The n2-node-128-864 sole-tenant node type.

Generally available: Support for the Intel Ice Lake processor on general purpose N2 VMs has reached general availability.

Rate limits for all Compute Engine requests have the following changes:

• All per-user rate limits are removed.
• Rate limits are now enforced in 1-minute (60-second) intervals instead of 100-second intervals.
• Due to this change, you might receive more 403 rateLimitExceeded errors when bursting.
  • Although per-second rate limits increased slightly, the enforcement intervals are now shorter, so the maximum number of requests per enforcement interval is slightly reduced overall. For example, the default Queries group's rate limit is changing from 20 requests per second with a maximum of 2000 requests per 100 seconds to 25 requests per second with a maximum of 1500 requests per 60 seconds.

Additionally, rate limits are now documented for the following groups:

• Instance list referrer requests
• Instance get serial port output requests

For details, see API rate limits.

Duplicate API quota groups are displayed in the Cloud Console. For more information about requesting API quota, see Known issues.

As of February 1, 2022, all CentOS 8 images are deprecated. CentOS 8 reached EOL on December 31, 2021. If you use CentOS 8 images in your project, review CentOS 8 end of life.

Restructured documentation to better group content and improve discoverability.

Generally available: Support for up to 48 vCPUs and 312 GB memory on virtual machine (VM) instances that have a single T4 GPU attached is now generally available.

For more information, see Network bandwidths and GPUs.

Learn about the differences between multi-tenancy and sole-tenancy by reading the new About VM tenancy document.

Generally available: You can now use the SSH troubleshooting tool to help you determine the cause of failed SSH connections.

Generally Available: Configure commitments to renew automatically. For more information, see Renew commitments automatically.

Generally available: Access the Compute Engine API using Cloud Client Libraries built on our latest client library model. Updated client libraries are now available in the following languages:

• Go
• Java
• .NET
• Node.js
• PHP
• Python
• Ruby

For more information, see Compute Engine client libraries.

Generally available: Compute Engine now supports machine images in General Availability. You can use machine images to store configuration, metadata, permission, and data required to create a VM instance.

Preview: You can now disable VM instance creation retries during resizing of both regional and zonal managed instance groups.

Accelerator-optimized (A2) machine types with gVNIC are currently experiencing a known issue.

Preview: Compute-optimized C2D machine types are now available in preview. C2D machine types are built on top of third generation AMD EPYC Milan processors and are a great fit for high-performance computing (HPC) workloads. For more information, see Compute-optimized machine family.

Generally available: When rolling out configuration or application updates to a stateful or stateless managed instance group, use the minimum and most disruptive allowed actions to control disruption to your workload.

Public preview: You can use the gcloud tool or API to configure stateful IP addresses in a managed instance group. Stateful IP addresses are preserved when VM instances in the group are autohealed, updated, and recreated.

Generally available: You can now share reservations of Compute Engine zonal resources between multiple projects. Learn about shared reservations and creating a shared reservation.

You can now save copies of all charts from the Observability tab on Compute Engine's VM instance details page to one of your custom dashboards. To save copies of the charts, click Add Charts to Dashboard. You then select a new or existing custom dashboard as the destination.

The n2-node-128-864 sole-tenant node type is now available in Preview.

Generally available: Use OS configuration management to deploy and automate software configurations on your virtual machine (VM) instances using the Google Cloud console, gcloud command-line, and OS Config API.

With the OS configuration management GA, you can now edit assignments from the Cloud console and view OS policy assignment reports. For more information, see OS configuration management.

Generally available: NVIDIA® A100 GPUs are now available in the following additional regions and zones:

• Moncks Corner, South Carolina : us-east1-b
• The Dalles, Oregon : us-west1-b
• Council Bluffs, Iowa : us-central1-f
• Jurong West, Singapore : asia-southeast1-b

For more information about using GPUs on Compute Engine, see GPUs on Compute Engine.

Generally available: Santiago, Chile, South America southamerica-west1-a,b,c region has launched with E2, N2, and C2 VMs in all three zones. See VM instance pricing for details.

Generally available: You can now configure N2, N2D, and C2 VMs with up to 100 Gbps of network bandwidth.

This feature is ideal for network-intensive, distributed workloads such as high-performance computing (HPC), machine learning (ML), and deep learning (DL).

Learn more about high-bandwidth network configurations, and the regions and zones where these VMs are available.

Generally available: T2D Tau machine types are available in select regions and zones. Tau T2D VMs offer excellent price-performance for a wide range of scale-out workloads. See VM instance pricing for details.

Generally available: N2D machine types running on third generation AMD EPYC Milan processors. These machine types are only available in specific regions and zones. See VM instance pricing for details.

Generally available: You can now monitor health state change logs for VM instances in a managed instance group when you use application-based health checking.

You can now access vulnerability report data, available through the OS Config API service, from Cloud Asset Inventory. For more information, see View vulnerability reports data from Cloud Asset Inventory.

Generally available: You can now use the gcloud command-line and the OS Config API to get inventory and vulnerability report data for your VMs in a specific zone. For more information, see Viewing operating system details.

If you use local SSDs with sync-heavy workloads, you will now more consistently reach write IOPS limits and experience lower latency, without having to disable cache flushing. This is due to a recent SSD firmware update.

You can now save a copy of a chart from the Observability tab on Compute Engine's VM instance details page to one of your custom dashboards. To save a copy of the chart, select Add to Custom Dashboard from chart option. You then select a new or existing custom dashboard, and have the option of renaming the new copy of the chart.

Generally available: Schedule-based autoscaling for managed instance groups now lets you configure schedules without having another autoscaling signal.

Preview: You can now configure up to 48 vCPUs and 312 GB memory on virtual machine (VM) instances that have a single T4 GPU attached.

For more information, see Network bandwidths and GPUs.

Preview: Spot VMs are now available! Spot VMs are the latest version of preemptible VM instances. Use Spot VMs for fault-tolerant workloads to get a 60-91% discount over the price of standard VMs. Spot prices can change up to once a month to reflect the underlying supply and demand. Like preemptible VMs, Spot VMs are available for all machine types, regions, and zones.

Preemptible VMs continue to be supported for new and existing VMs, and preemptible VMs now use the same pricing model as Spot VMs. However, Spot VMs provide new features that are not supported for preemptible VMs. For example, preemptible VMs can only run for up to 24 hours at a time, but Spot VMs do not have a maximum runtime.

Learn more about Spot VMs and preemptible VMs.

Preview: Third generation Intel Xeon Scalable Processor (Ice Lake) N2 VMs are now available in select regions and zones. These new N2 VMs are offered at the same price as existing N2 VMs on second generation Intel Xeon Scalable Processors.

Preview: Tau T2D VMs are now available in select regions and zones. T2D VMs are ideal for a wide variety of workloads in a cloud-native environment. See VM instance pricing for details.

Generally Available: Use patch alerting to monitor the patch jobs running in your environment. For more information, see Monitoring patch jobs.

Preview: You can now access installer properties for your Windows applications by using OS inventory management. For more information, see OS inventory management.

For information on setting up and using OS inventory management, see Viewing operating system details.

Generally Available: NVIDIA® T4 GPUs are now available in the following additional regions and zones:

• Las Vegas, Nevada,: us-west4-a,b
• Los Angeles, California: us-west2-b,c

For more information about using GPUs on Compute Engine, see GPUs on Compute Engine.

Preview: You can now review OS vulnerability report data, which is collected by VM Manager, from the Security Command Center. This feature is available for Security Command Center premium tier users. For more information, see View vulnerability report data.

Generally available: When deleting VMs from a managed instance group, you can flag the operation to continue even if some instances were already deleted or if other instance validation errors occur.

Generally available: You can now reference the latest available image in a public image family for a specific zone. This feature improves zonal fault tolerance for your workflows during Google image updates.

Generally available: You can now collect core dumps for uses such as debugging of unresponsive VMs. For more information, see Collecting core dumps.

Preview: Manually live migrate VMs from one host to another. For more information, see Manually live migrate sole-tenant VMs.

Preview: You can now use the Slurm-Google Cloud workload manager to create clusters that are based on the HPC virtual machine (VM) image and comply to the Intel Select Solution for Simulation and Modeling criteria. For more information, see Creating Intel Select Solution HPC clusters.

Generally available: The Observability tab on Compute Engine's VM instance details page includes a category for process metrics. You can use the new charts and reports to troubleshoot the behavior of processes running on your VMs.

N2 VMs are now available in all three zones in Warsaw, Poland europe-central2-a,b,c. See VM instance pricing for details.