create-service

Description

Create Service model.

A service is deployed based on specified version information then program is run n the container image environment specified in the version. Please wait minutes to access endpoint after execute a command cause it takes time to deploy.

Notes

Can not update or overwrite service. Please create new service and delete old one.

Service deployment method

You can create multiple services for deployment, and it is possible to do Blue Green Deployment such as switching the endpoint to a new service by specifying an optional service at update-endpoint.

Synopsis

$ abeja model create-service [--help]
Usage: abeja model create-service [OPTIONS]

  Create http service

Options:
  -d, --deployment_id, --deployment-id TEXT
                                  Deployment identifier  [required]
  -v, --version_id, --version-id TEXT
                                  Version identifier  [required]
  -m, --model_id, --model-id TEXT
                                  Model identifier
  -e, --environment ENVIRONMENTSTRING
                                  Service environment variable
  --instance-type, --instance_type TEXT
                                  Instance type
  --disable-autoscale, --disable_autoscale
                                  Disable autscale. Autoscale is enabled by
                                  default
  --instance-number, --instance_number INTEGER
                                  Number of intances. Available values are 1 to
                                  300. Default is 1
  --max-instance-number, --max_instance_number INTEGER
                                  Max number of intances of autoscaling.
                                  Default is double of instance-number
  --record-channel-id, --record_channel_id TEXT
                                  Channel identifier where input data will be
                                  saved
  --help                          Show this message and exit.

Options

-d, --deployment_id

Specify deployment_id with create-deployment you created

-v, --version_id

Specify version_id with create-deployment you created

-e, --environment

Specifies environment variables.The registered environment variables can be referenced from the code. E.g.）IMAGE_WIDTH:100
For more information on user-specifiable environment variables, see here.

--disable-autoscale, --disable_autoscale

Disable autoscale option of the service to create. If this option is not specified, autoscale is enabled.

--instance-number, --instance_number

Specifies the number of instances which run the service. Default values is 1.

--max-instance-number, --max_instance_number

Specifies the max number of instances which run the service. Default values is double of the --max-instance-number.

--instance-type, --instance_type

Specify instance type. you can select cpu-0.25, cpu-1, cpu-2, cpu-4 and default instance is cpu-0.25. ** gpu-1 is supported by CLI only. **

--record-channel-id, --record_channel_id

Specify channel ID to save the input data.

Example

Creating service

Command:

$ abeja model create-service --deployment_id 3456789012345     \
                             --version_id ver-2345678901234    \
                             --instance_type cpu-1             \
                             --record_channel_id 5678901234567

Output:

{
  "created_at": "2018-01-01T00:00:00.000000Z",
  "deployment_id": "3456789012345",
  "disruptable": true,
  "instance_number": 1,
  "instance_type": "cpu-1",
  "merged_env_vars": {},
  "metrics_url": "https://p.datadoghq.com/sb/xxxxxxxx",
  "model_version": "broken",
  "model_version_id": "ver-2345678901234",
  "modified_at": "2018-01-01T00:00:00.000000Z",
  "record_channel_id": "5678901234567",
  "service_id": "ser-4567890123456",
  "status": "IN_PROGRESS",
  "user_env_vars": {}
}