ModelDeployments provide an interface for tracking the health and activity of predictions
made against a deployment model. The GetModelDeploymentServiceStatistics method can
be used to see current and historical trends in requests made and in user and server
error rates.
GetModelDeployment(modelDeploymentId)character. ID of the model deployment to retrieve.
A dataRobotModelDeployment object containing the following data:
id character. The ID of the model deployment.
model list. Details on the model associated with the model deployment. Contains:
uid character. The ID of the user who created the model.
id character. The ID of the model.
modelType character. The type of the model.
prevRequestCount integer. The number of requests, within the previous time
window specified in trendTimeWindow.
project dataRobotProject. The project object associated with the model. See
GetProject for details.
type character. The type of the model deployment. One of "sse", "dedicated", or "legacy_dedicated".
status character. The status of the model deployment. One of "active", "inactive",
or "archived". Statuses can be accessed via ModelDeploymentStatus enum list.
user list. Details on the user who created the model deployment containing:
username character. The username of the user initiating deployment.
firstName character. The first name of that user.
lastName character. The last name of that user.
organizationId character. The ID of the organization associated with the model deployment.
instance list. Details on the instance associated with the model deployment. Contains the following information:
id character. The ID of the dedicated prediction instance the model is deployed to.
datarobotKey character. The key for the prediction instance.
hostName character. The host name of the dedicated prediction instance.
privateIp character. The IP address of the dedicated predicion instance.
ormVersion character. The On-demand resource manager version of the dedicated prediction instance.
label character. A short label describing the model deployment.
description character. A longer description describing the model deployment.
predictionEndpoint character. The URL where the model is deployed and available for serving predictions.
deployed logical. Whether the model deployment process has finished or not.
createdAt datetime. The timestamp of the creation of the model deployment.
updatedAt datetime. The timestamp of when the model deployment was last updated.
serviceHealth character. The model health status. One of "passing", "warning", or
"failing". Look at serviceHealthMessages or use
GetModelDeploymentServiceStatistics for more information.
serviceHealthMessages list. List of health messages detailing service health state. Contains the following information:
level character. The error level, one of "passing", "warning", or "failing".
msgId character. The identifier for the message, such as "USER_ERRORS", "SERVER_ERRORS", or "NO_GOOD_REQUESTS".
message character. A message describing the error with more detail.
recentRequestCount integer. The number of recent requests, within recent time window
specified by trendTimeWindow.
prevRequestCount integer. The number of requests within the previous time window
specified by trendTimeWindow.
relativeRequestsTrend numeric. Relative difference, as a percentage, between the
number of prediction requests performed within the current time window and one time
window before that. The size of the time window is specified by trendTimeWindow.
trendTimeWindow character. The time window, speficied in number of days from "now".
requestRates list. A history of request rates per day sorted in chronological order, with the last entry being the most recent (i.e., today).
# NOT RUN {
modelDeploymentId <- "59a5af20c80891534e3c2bde"
deployment <- GetTrainingPredictions(modelDeploymentId)
# }
Run the code above in your browser using DataLab