Skip to main content

anyscale

Check your docs version

This version of the Anyscale docs is deprecated. Go to the latest version for up to date information.

Options:

  • -v, --version Current anyscale version.
  • --json Return output as json, for use with --version.
  • --help Show this message and exit.

Commands:

  • cloud Configure cloud provider authentication for Anyscale.
  • cluster Interact with clusters on Anyscale.
  • cluster-env Interact with cluster environments.
  • compute-config Manage compute configurations on Anyscale.
  • fine-tuning Interact with fine-tuning jobs running on Anyscale.
  • job Interact with production jobs running on Anyscale.
  • login Log in to Anyscale using a URL.
  • logout Log out from Anyscale.
  • logs Print or download Ray logs for an Anyscale job,...
  • machine Commands to interact with machines in Anyscale.
  • machine-pool Commands to interact with machine pools in Anyscale.
  • migrate Manages configurations such as cluster environments and...
  • project Manage projects on Anyscale.
  • schedule Create and manage Anyscale Schedules.
  • service Interact with production services running on Anyscale.
  • service-account Manage service accounts on Anyscale.
  • version Display version of the anyscale CLI.
  • workspace Interact with workspaces on Anyscale.

anyscale cloud​

Configure cloud provider authentication and setup to allow Anyscale to launch instances in your account.

Options:

  • --help Show this message and exit.

Commands:

  • config Manage the configuration for a cloud.
  • delete Delete a cloud.
  • edit Edit anyscale cloud resource.
  • list List information about clouds in your Anyscale organization.
  • register Register an anyscale cloud with your own resources.
  • set-default Sets default cloud for your organization.
  • setup Set up a cloud provider.
  • update Update managed cloud.
  • verify Checks the healthiness of a cloud.

anyscale cloud config​

Manage the configuration for a cloud.

Options:

  • --help Show this message and exit.

Commands:

  • get Get the current configuration for a cloud.

anyscale cloud config get​

Get the current configuration for a cloud.

Options:

  • -n, --name TEXT Get configuration of cloud by name.
  • --cloud-id, --id TEXT Cloud id to get details about. Alternative to cloud name.
  • --help Show this message and exit.

anyscale cloud config update​

Update the current configuration for a cloud.

Options:

  • -n, --name TEXT Update configuration of cloud by name.
  • --cloud-id, --id TEXT Cloud id to get details about. Alternative to cloud name.
  • --enable-log-ingestion / --disable-log-ingestion If --enable-log-ingestion is specified for a cloud, it will enable the log viewing and querying UI features for the clusters on this cloud. This will enable easier debugging. The logs produced by the clusters will be sent from the data plane to the control plane. Anyscale does not share this data with any third party or use it for any purpose other than serving the log UI for the customer. The log will be stored at most 30 days. Please note by disable this feature again, Anyscale doesn't delete the logs that have already been ingested. Your clusters may incur extra data transfer cost from the cloud provider by enabling this feature.
  • --help Show this message and exit.

anyscale cloud delete​

Delete a cloud.

Options:

  • -n, --name TEXT Delete cloud by name.
  • --cloud-id, --id TEXT Cloud id to delete. Alternative to cloud name.
  • -y, --yes Don't ask for confirmation.
  • --help Show this message and exit.

anyscale cloud edit​

Edit anyscale cloud resource.

Options:

  • -n, --name TEXT Edit cloud by name.
  • --cloud-id, --id TEXT Edit cloud by id, alternative to cloud name.
  • --aws-s3-id TEXT New S3 bucket ID.
  • --aws-efs-id TEXT New EFS ID.
  • --aws-efs-mount-target-ip TEXT New EFS mount target IP.
  • --memorydb-cluster-id TEXT New AWS Memorydb cluster ID.
  • --gcp-filestore-instance-id TEXT New GCP filestore instance id.
  • --gcp-filestore-location TEXT New GCP filestore location.
  • --gcp-cloud-storage-bucket-name TEXT New GCP Cloud storage bucket name.
  • --memorystore-instance-name TEXT New Memorystore instance name for GCP clouds
  • --functional-verify TEXT Verify the cloud is functional. This will check that the cloud can launch workspace/service.
  • --enable-auto-add-user / --disable-auto-add-user If --enable-auto-add-user is specified for a cloud, all users in the organization will be added to the cloud by default. Note: There may be up to 30 sec delay for all users to be granted permissions after this feature is enabled. Specifying --disable-auto-add-user will require that users are manually granted permissions to access the cloud. No existing cloud permissions are altered by specifying this flag.
  • --help Show this message and exit.

anyscale cloud list​

List information about clouds in your Anyscale organization.

Options:

  • -n, --name TEXT Name of cloud to get information about.
  • --cloud-id, --id TEXT Id of cloud to get information about.
  • --max-items INTEGER Max items to show in list.
  • --help Show this message and exit.

anyscale cloud register​

Register an anyscale cloud with your own resources.

Options:

  • --provider [aws|gcp] The cloud provider type. [required]
  • --region TEXT Region to set up the credentials in. [required]
  • -n, --name TEXT Name of the cloud. [required]
  • --vpc-id TEXT The ID of the VPC.
  • --subnet-ids TEXT Comma separated list of subnet ids.
  • --efs-id TEXT The EFS ID.
  • --anyscale-iam-role-id TEXT The Anyscale IAM Role ARN.
  • --instance-iam-role-id TEXT The instance IAM role ARN.
  • --security-group-ids TEXT IDs of the security groups.
  • --s3-bucket-id TEXT S3 bucket ID.
  • --memorydb-cluster-id TEXT Memorydb cluster ID
  • --project-id TEXT Globally Unique project ID for GCP clouds (e.g., my-project-abc123)
  • --vpc-name TEXT VPC name for GCP clouds
  • --subnet-names TEXT Comma separated list of subnet names for GCP clouds
  • --filestore-instance-id TEXT Filestore instance ID for GCP clouds.
  • --filestore-location TEXT Filestore location for GCP clouds.
  • --anyscale-service-account-email TEXT Anyscale service account email for GCP clouds.
  • --instance-service-account-email TEXT Instance service account email for GCP clouds.
  • --provider-name TEXT Workload Identity Federation provider name for Anyscale access.
  • --firewall-policy-names TEXT Filewall policy names for GCP clouds
  • --cloud-storage-bucket-name TEXT Cloud storage bucket name for GCP clouds
  • --memorystore-instance-name TEXT Memorystore instance name for GCP clouds
  • --host-project-id TEXT Host project ID for shared VPC
  • --private-network Use private network.
  • --functional-verify TEXT Verify the cloud is functional. This will check that the cloud can launch workspace/service.
  • -y, --yes Skip asking for confirmation.
  • --skip-verifications Skip verifications. This will skip all verifications.
  • --enable-auto-add-user If --enable-auto-add-user is specified for a cloud, all users in the organization will be added to the cloud by default. Otherwise users will need to be manually granted permissions to the cloud. Note: There may be up to 30 sec delay for all users to be granted permissions after the cloud is created.
  • --help Show this message and exit.

anyscale cloud set-default​

Sets default cloud for your organization. This operation can only be performed by organization admins, and the default cloud must have organization level permissions.

Options:

  • -n, --name TEXT Set cloud as default by name.
  • --cloud-id, --id TEXT Cloud id to set as default. Alternative to cloud name.
  • --help Show this message and exit.

anyscale cloud setup​

Set up a cloud provider.

Options:

  • --provider [aws|gcp] The cloud provider type. [required]
  • --region TEXT Region to set up the credentials in. [required]
  • -n, --name TEXT Name of the cloud. [required]
  • --project-id TEXT Globally Unique project ID for GCP clouds (e.g., my-project-abc123)
  • --functional-verify TEXT Verify the cloud is functional. This will check that the cloud can launch workspace/service.
  • --anyscale-managed Let anyscale create all the resources.
  • --enable-head-node-fault-tolerance Whether to enable head node fault tolerance for services.
  • -y, --yes Skip asking for confirmation.
  • --disable-auto-add-user All users in the organization will be added to clouds created with anyscale cloud setup by default. Specify --disable-auto- add-user to disable this and instead manually grant users permissions to the cloud.
  • --help Show this message and exit.

anyscale cloud update​

Update managed cloud.

Options:

  • --cloud-id, --id TEXT Cloud id to update. Alternative to cloud name.
  • -n, --name TEXT Update configuration of cloud by name.
  • -y, --yes Skip asking for confirmation.
  • --functional-verify TEXT Verify the cloud is functional. This will check that the cloud can launch workspace/service.
  • --enable-head-node-fault-tolerance Whether to enable head node fault tolerance for services.
  • --enable-auto-add-user / --disable-auto-add-user If --enable-auto-add-user is specified for a cloud, all users in the organization will be added to the cloud by default. Note: There may be up to 30 sec delay for all users to be granted permissions after this feature is enabled. Specifying --disable-auto-add-user will require that users are manually granted permissions to access the cloud. No existing cloud permissions are altered by specifying this flag.
  • --help Show this message and exit.

anyscale cloud verify​

Checks the healthiness of a cloud.

Options:

  • -n, --name TEXT Verify cloud by name.
  • --cloud-id, --id TEXT Verify cloud by cloud id, alternative to cloud name.
  • --functional-verify TEXT Verify the cloud is functional. This will check that the cloud can launch workspace/service.
  • --strict Strict Verify. Treat warnings as failures.
  • --help Show this message and exit.

anyscale cluster​

Interact with clusters on Anyscale.

Options:

  • --help Show this message and exit.

Commands:

  • archive Archive a cluster on Anyscale.
  • list List information about clusters on Anyscale.
  • start Start or update and restart a cluster on Anyscale.
  • terminate Terminate a cluster on Anyscale.

anyscale cluster archive​

Archive a cluster on Anyscale.

Options:

  • -n, --name TEXT Name of existing cluster to archive.
  • --cluster-id, --id TEXT Id of existing cluster to archive. This argument can be used to archive any cluster you have access to in any project.
  • --project-id TEXT Override project id used for this cluster. If not provided, the Anyscale project context will be used if it exists. Otherwise a default project will be used.
  • --project TEXT Override project name used for this cluster. If not provided, the Anyscale project context will be used if it exists. Otherwise a default project will be used.
  • --help Show this message and exit.

anyscale cluster list​

List information about clusters on Anyscale. By default only list active clusters in current project.

Options:

  • -n, --name TEXT Name of existing cluster to get information about.
  • --cluster-id, --id TEXT Id of existing cluster get information about. This argument can be used to interact with any cluster you have access to in any project.
  • --project-id TEXT Override project id used for this cluster. If not provided, the Anyscale project context will be used if it exists. Otherwise a default project will be used.
  • --project TEXT Override project name used for this cluster. If not provided, the Anyscale project context will be used if it exists. Otherwise a default project will be used.
  • --include-all-projects List all active clusters user has access to in any project.
  • --include-inactive List clusters of all states.
  • --include-archived List archived clusters as well as unarchived clusters.If not provided, defaults to listing only unarchived clusters.
  • --max-items INTEGER Max items to show in list.
  • --help Show this message and exit.

anyscale cluster start​

Start or update and restart a cluster on Anyscale.

Options:

  • -n, --name TEXT Name of new or existing cluster to start or update.
  • --env TEXT Set the Anyscale app config to use for the cluster. This is a cluster environment name optionally followed by a colon and a build version number. Eg: my_cluster_env:1
  • --compute TEXT Name of cluster compute that is already registered with Anyscale.
  • --compute-file TEXT The YAML file of the cluster compute config to launch this cluster with. An example can be found at https://docs.anyscale.com/configure/compute- configs/overview.
  • --cluster-id, --id TEXT Id of existing cluster to restart. This argument can be used to interact with any cluster you have access to in any project.
  • --project-id TEXT Override project id used for this cluster. If not provided, the Anyscale project context will be used if it exists. Otherwise a default project will be used.
  • --project TEXT Override project name used for this cluster. If not provided, the Anyscale project context will be used if it exists. Otherwise a default project will be used.
  • --cloud-name TEXT Name of cloud to create a default cluster compute with. If a default cloud needs to be used and this is not provided, the organization default cloud will be used.
  • --idle-timeout INTEGER DEPRECATED: Please specify the idle_termination_minutes field in the cluster compute. Idle timeout (in minutes), after which the cluster is stopped. Idle time is defined as the time during which a cluster is not running a user command and does not have an attached driver. Time spent running Jupyter commands, or commands run through ssh, is still considered 'idle'. -1 means no timeout. Default: 120 minutes
  • --user-service-access [private|public] Whether user service (eg: serve deployment) can be accessed by public internet traffic. If public, a user service endpoint can be queried from the public internet with the provided authentication token. If private, the user service endpoint can only be queried from within the same Anyscale cloud and will not require an authentication token.
  • --cluster-compute-version INTEGER The cluster compute version that you want to use. Only applicable if the compute was specified by name. Versions are positive integers. Deprecated behavior: Not setting a version is equivalent to setting version to '1'.
  • --help Show this message and exit.

anyscale cluster terminate​

Terminate a cluster on Anyscale.

Options:

  • -n, --name TEXT Name of existing cluster to terminate.
  • --cluster-id, --id TEXT Id of existing cluster to termiante. This argument can be used to interact with any cluster you have access to in any project.
  • --project-id TEXT Override project id used for this cluster. If not provided, the Anyscale project context will be used if it exists. Otherwise a default project will be used.
  • --project TEXT Override project name used for this cluster. If not provided, the Anyscale project context will be used if it exists. Otherwise a default project will be used.
  • --help Show this message and exit.

anyscale cluster-env​

Interact with cluster environments.

Options:

  • --help Show this message and exit.

Commands:

  • archive Archive the specified cluster environment.
  • build Build a new cluster environment from config file.
  • get Get details about cluster environment build.
  • list List information about cluster environments on Anyscale.

anyscale cluster-env archive​

Archive the specified cluster environment.

Options:

  • -n, --name TEXT Name of the cluster environment to archive.
  • --cluster-env-id, --id TEXT Id of the cluster environment to archive. Must be provided if a cluster environment name is not given.
  • --help Show this message and exit.

anyscale cluster-env build​

Build a new cluster environment from config file.

Options:

  • -n, --name TEXT Name to save built cluster environment as. Default will be used if not provided
  • --help Show this message and exit.

anyscale cluster-env get​

Get details about cluster environment build. The cluster-env-name argument is a cluster environment name optionally followed by a colon and a build version number. Eg: my_cluster_env:1

Options:

  • --cluster-env-build-id, --id TEXT Get details about cluster environment build by this id.
  • --help Show this message and exit.

anyscale cluster-env list​

List information about cluster environments on Anyscale. By default only list cluster environments you have created.

Options:

  • -n, --name TEXT List information about all builds of the cluster environment with this name.
  • --cluster-env-id, --id TEXT List information about all builds of the cluster environment with this id.
  • --include-shared Include all cluster environments you have access to.
  • --max-items INTEGER Max items to show in list.
  • --help Show this message and exit.

anyscale compute-config​

Manages compute configurations to define cloud resource types and limitations.

Options:

  • --help Show this message and exit.

Commands:

  • archive Archive the specified compute config.
  • create Creates a new compute config.
  • get Get details about compute configuration.
  • list List information about compute configs on Anyscale.

anyscale compute-config archive​

Archive the specified compute config.

Options:

  • -n, --name TEXT Name of the compute config to archive.
  • --cluster-compute-id, --id TEXT Id of the compute config to archive. Must be provided if a compute name is not given.
  • --help Show this message and exit.

anyscale compute-config create​

Creates a new compute config. This accepts a yaml that follows the schema defined at https://docs.anyscale.com/reference/python- sdk/models#createclustercomputeconfig

Options:

  • -n, --name TEXT Name for the created compute config.
  • --help Show this message and exit.

anyscale compute-config get​

Get details about compute configuration.

Options:

  • --cluster-compute-id, --id TEXT Get details about compute configuration by this id.
  • --help Show this message and exit.

anyscale compute-config list​

List information about compute configs on Anyscale. By default only list compute configs you have created.

Options:

  • -n, --name TEXT List information about the compute config with this name.
  • --cluster-compute-id, --id TEXT List information about the compute config with this id.
  • --include-shared Include all compute configs you have access to.
  • --max-items INTEGER Max items to show in list.
  • --help Show this message and exit.

anyscale fine-tuning​

Interact with fine-tuning jobs running on Anyscale.

Options:

  • --help Show this message and exit.

Commands:

  • submit Submits a fine-tuning job.

anyscale fine-tuning submit​

Submits a fine-tuning job.

Example usage:

anyscale fine-tuning submit --model meta-llama/Llama-2-7b-chat-hf --train-file train.jsonl --cloud-id CLOUD_ID

Options:

  • --train-file TEXT The path of the training file. [required]
  • --valid-file TEXT The path of the validation file.
  • --cloud-id TEXT The id of the cloud [required]
  • --suffix TEXT The suffix of the fine-tuned model.
  • --version TEXT The version of ray-llm expected to be used for the fine-tuned model. If not specified, the latest version will be used.
  • --instance-type TEXT The instance type to use for the fine-tuning job. Please refer to docs at https://docs.anyscale.com/endpoints/fine-tuning/get- started for available instance types.If not specified, the default instance type will be used.
  • --help Show this message and exit.

anyscale job​

Interact with production jobs running on Anyscale.

Options:

  • --help Show this message and exit.

Commands:

  • archive Archive a job.
  • list Display information about existing jobs.
  • logs Print the logs of a job.
  • output DEPRECATED: Interact with Production Job outputs.
  • submit Submit a Production Job
  • terminate Attempt to terminate a job asynchronously.
  • wait Wait for a Job to enter a specific state (default: SUCCESS).

anyscale job archive​

Archive a job.

Options:

  • --job-id, --id TEXT Unique ID of the job.
  • -n, --name TEXT Name of the job.
  • --help Show this message and exit.

anyscale job list​

Display information about existing jobs.

Options:

  • -n, --name TEXT Filter by job name.
  • --job-id, --id TEXT Filter by job id.
  • --project-id TEXT Filter by project id.
  • --include-all-users Include jobs not created by current user.
  • --include-archived List archived jobs as well as unarchived jobs.If not provided, defaults to listing only unarchived jobs.
  • --max-items INTEGER Max items to show in list.
  • --help Show this message and exit.

anyscale job logs​

Print the logs of a job.

By default from the latest job attempt.

Example usage:

anyscale job logs --id prodjob_123

anyscale job logs --id prodjob_123 -f

anyscale job logs -n my-job --all-attempts

Options:

  • --job-id, --id TEXT Unique ID of the job.
  • -n, --name TEXT Name of the job.
  • -f, --follow Whether to follow the log.
  • --all-attempts Show logs for all job attempts.
  • --help Show this message and exit.

anyscale job output​

DEPRECATED: Interact with Production Job outputs. Job outputs do not work in V2 stack and has been removed from the UI. The CLI commands will be removed in the future.

Options:

  • --help Show this message and exit.

Commands:

  • get Get the output of a Production Job.
  • write Write the output of a Production Job to Anyscale.

anyscale job output get​

Get the output of a Production Job.

Options:

  • --job-id, --id TEXT Unique ID of the job.
  • -n, --name TEXT Name of the job.
  • --help Show this message and exit.

anyscale job output write​

Write the output of a Production Job to Anyscale.

This command should be called from within a Production Job entrypoint. It will record the JSON output of your Job. If you specify neither a file or a string, this command will read from STDIN.

Examples:

anyscale job output write -f output.json

anyscale job output write '{"output": 5}'

cat output.json | anyscale job output write

Options:

  • -f, --file FILENAME File to read the output from.
  • --help Show this message and exit.

anyscale job submit​

Submit a Production Job

This function accepts 1 argument, a path to a YAML config file that defines this job.

To wait for the Job to complete, use the --wait flag as follows.

$ anyscale job submit job.yaml --wait

If running from an Anyscale Workspace, this command also support an alternative syntax to specify the entrypoint inline:

$ anyscale job submit -- python script.py

The Cluster Environment and Cluster Config are inferred from the Workspace Configuration if running from a workspace.

Options:

  • -n, --name TEXT Name of the job.
  • --description TEXT Description of job.
  • -f, --follow Whether to follow the log of the created job
  • -w, --wait Whether to block this CLI command until the job succeeds (or reaches another terminal state)
  • --help Show this message and exit.

anyscale job terminate​

Attempt to terminate a job asynchronously.

Options:

  • --job-id, --id TEXT Unique ID of the job.
  • -n, --name TEXT Name of the job.
  • --help Show this message and exit.

anyscale job wait​

Wait for a Job to enter a specific state (default: SUCCESS).

To specify the Job by name, use the --name flag. To specify the Job by id, use the --job-id flag.

If the Job reaches the target state, the command will exit successfully.

If the Job reaches a terminal state other than the target state, the command will exit with an error.

If the command reaches the timeout, the command will exit with an error.

Options:

  • --job-id, --id TEXT Unique ID of the job.
  • -n, --name TEXT Name of the job.
  • -s, --state TEXT The state to wait for this Job to enter
  • -t, --timeout FLOAT The timeout in seconds after which this command will exit.
  • --help Show this message and exit.

anyscale login​

Log in to Anyscale using a URL.

Options:

  • --no-expire Do not expire the token.
  • --expire-in-days INTEGER Expire the token after this many days.
  • --help Show this message and exit.

anyscale logout​

Log out from Anyscale. You CLI token will be revoked.

Options:

  • --help Show this message and exit.

anyscale logs​

Print or download Ray logs for an Anyscale job, service, or cluster.

Options:

  • --help Show this message and exit.

Commands:

  • cluster Access log files of a cluster.
  • job Access log files of a production job.

anyscale logs cluster​

Access log files of a cluster.

Options:

  • --id TEXT Provide a cluster ID. [required]
  • -d, --download Download logs to the current working directory, or a specified path.
  • -t, --tail INTEGER Read the last N lines of logs.
  • -ip, --node-ip TEXT Filter logs by a node IP.
  • --instance-id TEXT Filter logs by an instance ID.
  • --worker-only Download logs of only the worker nodes.
  • --head-only Download logs of only the head node.
  • --download-dir TEXT Directory to download logs into.
  • --help Show this message and exit.

anyscale logs job​

Access log files of a production job.

Options:

  • --id TEXT Provide a production job ID. [required]
  • -d, --download Download logs to the current working directory, or a specified path.
  • -t, --tail INTEGER Read the last N lines of logs.
  • -ip, --node-ip TEXT Filter logs by a node IP.
  • --instance-id TEXT Filter logs by an instance ID.
  • --worker-only Download logs of only the worker nodes.
  • --head-only Download logs of only the head node.
  • --download-dir TEXT Directory to download logs into.
  • --help Show this message and exit.

anyscale machine​

Commands to interact with machines in Anyscale.

Options:

  • --help Show this message and exit.

Commands:

  • delete Delete a machine from a machine pool.
  • list List machines registered to Anyscale.

anyscale machine delete​

Delete a machine from a machine pool.

Options:

  • --cloud TEXT Provide a cloud name.
  • --cloud-id TEXT Provide a cloud ID.
  • --machine-id TEXT Provide a machine ID.
  • --help Show this message and exit.

anyscale machine list​

List machines registered to Anyscale.

Options:

  • --cloud TEXT Provide a cloud name.
  • --cloud-id TEXT Provide a cloud ID.
  • --machine-pool-name TEXT Provide a machine pool name.
  • --format [table|json] The format to return results in.
  • --help Show this message and exit.

anyscale machine-pool​

Commands to interact with machine pools in Anyscale.

Options:

  • --help Show this message and exit.

Commands:

  • attach Attach a machine pool to a cloud.
  • create Create a machine pool in Anyscale.
  • delete Delete a machine pool in Anyscale.
  • detach Detach a machine pool from a cloud.
  • list List machine pools in Anyscale.

anyscale machine-pool attach​

Attach a machine pool to a cloud.

Options:

  • --name TEXT Provide a machine pool name. [required]
  • --cloud TEXT Provide a cloud name. [required]
  • --help Show this message and exit.

anyscale machine-pool create​

Create a machine pool in Anyscale.

Options:

  • --name TEXT Provide a machine pool name (must be unique within an organization). [required]
  • --help Show this message and exit.

anyscale machine-pool delete​

Delete a machine pool in Anyscale.

Options:

  • --name TEXT Provide a machine pool name. [required]
  • --help Show this message and exit.

anyscale machine-pool detach​

Detach a machine pool from a cloud.

Options:

  • --name TEXT Provide a machine pool name. [required]
  • --cloud TEXT Provide a cloud name. [required]
  • --help Show this message and exit.

anyscale machine-pool list​

List machine pools in Anyscale.

Options:

  • --help Show this message and exit.

anyscale migrate​

Manages configurations such as cluster environments and compute configs that help you run jobs and clusters on Anyscale.

Options:

  • --help Show this message and exit.

Commands:

  • convert Converts a Ray cluster YAML into a cluster environment...
  • upload-configs Uploads both the cluster-env and compute-config.

anyscale migrate convert​

Converts a Ray cluster YAML into a cluster environment and compute config.

Options:

  • --cloud TEXT Cloud to create compute_config for.
  • --ml / --no-ml Whether to generate a cluster-env that contains ray-ml dependencies
  • --gpu / --no-gpu Whether to generate a cluster-env that contains the required dependencies to utilize a gpu
  • --help Show this message and exit.

anyscale migrate upload-configs​

Uploads both the cluster-env and compute-config.

Options:

  • -n, --name TEXT Name for both configs [required]
  • --help Show this message and exit.

anyscale project​

Manages projects on Anyscale. A project can be used to organize a collection of jobs.

Options:

  • --help Show this message and exit.

Commands:

  • create Create a new project.
  • list List projects for which you have access.

anyscale project create​

Create a new project.

Options:

  • -n, --name TEXT Project name.
  • --parent-cloud-id TEXT Cloud id that this project is associated with. This argument is only relevant if cloud isolation is enabled.
  • --help Show this message and exit.

anyscale project list​

List projects for which you have access. By default, only projects created by you are listed.

Options:

  • -n, --name TEXT List information for a particular project.
  • --json Format output as JSON.
  • --created-by-me List projects created by me only.
  • --max-items INTEGER Max items to show in list.
  • --help Show this message and exit.

anyscale schedule​

Create and manage Anyscale Schedules.

Options:

  • --help Show this message and exit.

Commands:

  • create Create or Update a Schedule
  • list List Schedules
  • pause Pause a Schedule
  • resume Resume a Schedule
  • run Manually run a Schedule
  • update Create or Update a Schedule
  • url Get a Schedule URL

anyscale schedule create​

Create or Update a Schedule

This function accepts 1 argument, a path to a YAML config file that defines this schedule.

Note: if a schedule with the name exists in the specified project, it will be updated instead.

Options:

  • -n, --name TEXT Name of the schedule.
  • --description TEXT Description of schedule.
  • --help Show this message and exit.

anyscale schedule list​

List Schedules

You can optionally filter schedules by rowname.

Options:

  • -n, --name TEXT Filter by the name of the job
  • -i, --id TEXT Id of the schedule.
  • --help Show this message and exit.

anyscale schedule pause​

Pause a Schedule

This function accepts 1 argument, a path to a YAML config file that defines this schedule. You can also specify the schedule by name or id.

Options:

  • -n, --name TEXT Name of the schedule.
  • -i, --id TEXT Id of the schedule.
  • --help Show this message and exit.

anyscale schedule resume​

Resume a Schedule

This function accepts 1 argument, a path to a YAML config file that defines this schedule. You can also specify the schedule by name or id.

Options:

  • -n, --name TEXT Name of the schedule.
  • -i, --id TEXT Id of the schedule.
  • --help Show this message and exit.

anyscale schedule run​

Manually run a Schedule

This function takes an existing schedule and runs it now. You can specify the schedule by name or id. You can also pass in a YAML file as a convinience. This is equivalent to passing in the name specified in the YAML file. IMPORTANT: if you pass in a YAML definition that differs from the Schedule defition, the Schedule will NOT be updated. Please use the anyscale schedule update command to update the configuration of your schedule or use the anyscale job submit command to submit a one off job that is not a part of a schedule.

Options:

  • -n, --name TEXT Name of the schedule.
  • -i, --id TEXT Id of the schedule.
  • --help Show this message and exit.

anyscale schedule update​

Create or Update a Schedule

This function accepts 1 argument, a path to a YAML config file that defines this schedule.

Options:

  • -n, --name TEXT Name of the schedule.
  • --description TEXT Description of schedule.
  • --help Show this message and exit.

anyscale schedule url​

Get a Schedule URL

This function accepts 1 argument, a path to a YAML config file that defines this schedule. You can also specify the schedule by name or id.

Options:

  • -n, --name TEXT Name of the schedule.
  • -i, --id TEXT Id of the schedule.
  • --help Show this message and exit.

anyscale service​

Interact with production services running on Anyscale.

Options:

  • --help Show this message and exit.

Commands:

  • list Display information about existing services.
  • rollback Attempt to rollback a v2 service asynchronously.
  • rollout Rollout a v2 service to Anyscale.
  • terminate Attempt to terminate a service asynchronously.

anyscale service list​

Display information about existing services.

Options:

  • -n, --name TEXT Filter by service name.
  • --service-id, --id TEXT Filter by service id.
  • --project-id TEXT Filter by project id.
  • --created-by-me List services created by me only.
  • --max-items INTEGER Max items to show in list.
  • --help Show this message and exit.

anyscale service rollback​

Attempt to rollback a v2 service asynchronously.

Options:

  • --service-id, --id TEXT ID of service.
  • -n, --name TEXT Name of service.
  • --project-id TEXT Filter by project id.
  • -f, --service-config-file TEXT The path of the service configuration file.
  • --max-surge-percent INTEGER Max amount of excess capacity allocated during the rollback (0-100).
  • --help Show this message and exit.

anyscale service rollout​

Rollout a v2 service to Anyscale. Please contact support for details.

Options:

  • -f, --service-config-file TEXT The path of the service configuration file. [required]
  • -n, --name TEXT Name of service.
  • --version TEXT Version of service.
  • --max-surge-percent INTEGER Max amount of excess capacity allocated during the rollout (0-100).
  • --canary-percent INTEGER The percentage of traffic to send to the canary version of the service.
  • --rollout-strategy [ROLLOUT|IN_PLACE] Strategy for rollout.
  • -i, --in-place Alias for --rollout-strategy=IN_PLACE. [default: False]
  • --no-auto-complete-rollout Do not complete the rollout (terminate the existing version cluster) after the canary version reaches 100% [default: False]
  • --help Show this message and exit.

anyscale service terminate​

Attempt to terminate a service asynchronously.

Options:

  • --service-id, --id TEXT ID of service.
  • -n, --name TEXT Name of service.
  • --project-id TEXT Filter by project id.
  • -f, --service-config-file TEXT The path of the service configuration file.
  • --help Show this message and exit.

anyscale service-account​

anyscale service-account create​

Usage

anyscale service-account create [OPTIONS]

Create a service account.

Options

  • --name/-n: Name for the service account.

anyscale service-account create-api-key​

Usage

anyscale service-account create-api-key [OPTIONS]

Create a new API key for a service account.

Options

  • --email: Email of the service account to create the new key for.
  • --name: Name of the service account to create the new key for.

anyscale service-account list​

Usage

anyscale service-account list [OPTIONS]

List all service accounts.

Options

  • --max-items: Max items to show in list.

anyscale service-account delete​

Usage

anyscale service-account delete [OPTIONS]

Delete a service account.

Options

  • --email: Email of the service account to delete.
  • --name: Name of the service account to delete.

anyscale service-account rotate-api-keys​

Usage

anyscale service-account rotate-api-keys [OPTIONS]

Rotate all API keys of a service account.

Options

  • --email: Rotate API keys for the service account with this email.
  • --name: Rotate API keys for the service account with this name.

anyscale version​

Display version of the anyscale CLI.

Options:

  • --json Return output as json.
  • --help Show this message and exit.

anyscale workspace​

Interact with workspaces on Anyscale.

Options:

  • --help Show this message and exit.

Commands:

  • activate Activate a workspace.
  • clone Clone a workspace on Anyscale.
  • cp Copy a file or a directory from workspace to local file system.
  • create Create a workspace on Anyscale.
  • pull Pull files from a workspace on Anyscale.
  • push Push files to a workspace on Anyscale.
  • run Run a command in a workspace, syncing files first if needed.
  • ssh ssh into a workspace, you can also pass args to the ssh...
  • start Start an existing workspace on Anyscale.
  • terminate Terminate a workspace on Anyscale.

anyscale workspace activate​

Activate a workspace.

If the current directory is already a part of a workspace, change the workspace. Else, setup a new workspace rooted at the current directory

Args: name: Name of the workspace to activate.

Options:

  • --help Show this message and exit.

anyscale workspace clone​

Clone a workspace on Anyscale.

Options:

  • -n, --name TEXT Name of existing workspace to clone. [required]
  • --help Show this message and exit.

anyscale workspace cp​

Copy a file or a directory from workspace to local file system.

Examples

anyscale workspace cp /mnt/shared_objects/foo.py /tmp/copy_of_foo.py

anyscale workspace cp "~/default/README.md" ~/Downlaods

anyscale workspace cp "/tmp/" ~/Downlaods

Options:

  • --help Show this message and exit.

anyscale workspace create​

Create a workspace on Anyscale.

Options:

  • -n, --name TEXT Name of the workspace to create. [required]
  • --project-id TEXT [required]
  • --cloud-id TEXT [required]
  • --cluster-env-build-id TEXT
  • --docker TEXT Custom docker image URI.
  • --python-version TEXT Python version for the custom docker image.
  • --ray-version TEXT Ray version for the custom docker image.
  • --compute-config-id TEXT [required]
  • --help Show this message and exit.

anyscale workspace pull​

Pull files from a workspace on Anyscale.

Options:

  • --pull-git-state Also pull git state. This will add additional overhead.
  • --help Show this message and exit.

anyscale workspace push​

Push files to a workspace on Anyscale.

Options:

  • --push-git-state Also push git state. This is currently unoptimized and will be very slow.
  • --help Show this message and exit.

anyscale workspace run​

Run a command in a workspace, syncing files first if needed.

Options:

  • -w, --web-terminal Run the command in the webterminal. Progress can be tracked from the UI.
  • -j, --as-job Run the command as a background job in a new cluster.
  • -s, --no-push Whether to skip pushing files prior to running the command.
  • --help Show this message and exit.

anyscale workspace ssh​

ssh into a workspace, you can also pass args to the ssh command. E.g. 'anyscale workspace ssh -- -L 8888:localhost:8888

Options:

  • --help Show this message and exit.

anyscale workspace start​

Start an existing workspace on Anyscale.

Options:

  • -n, --name TEXT Name of existing workspace to start. [required]
  • --help Show this message and exit.

anyscale workspace terminate​

Terminate a workspace on Anyscale.

Options:

  • -n, --name TEXT Name of existing workspace to terminate. [required]
  • --help Show this message and exit.