System audit log

The audit log for the system records the import and remove actions of models and deployments. Use the following steps to view the audit log for the system.

Sign in to Minitab Model Ops.
Select System Audit Log .

Note

The audit log for each deployment is found on the Audit Log tab for the deployment.

Audit log information

The log records the following items for each action. Expand an item to see additional details for the action.

Timestamp

The date and time of the action.

User

The name from the User Information for the account for Minitab Model Ops.

Action

This value records one of the following actions.

ImportModel
UpdateModel
RemoveModel
CreateDeployment
RemoveDeployment
UploadDataset

Result

This value records whether the action succeeds or fails.

Note

Audit logs entries older than 1 year are purged from the system.

Import model failures

The system audit log records the following failures when importing a model.

InvalidColumns: The model import fails if you use a reserved variable name. For more information, go to Reserved variable names.
InvalidSchemaText: The model import fails if you use an invalid character in the response or class level. For example, the characters that represent a new line of '\n' and '\r' cannot be used.
TooManyNodes: The model is too large. Reduce the number of trees and/or the node size for your model.
TooManyPredictors: Models must have 500 predictors or less.
TooManyResponseClasses: The maximum number of classes is 10 for a multinomial response variable.

Import baseline data failures

The system audit log records the following failures when importing baseline data for a model.

BaselineFileSizeLimit: The baseline data import fails if the file size is larger than 10 megabytes.
InvalidDataset: The baseline data file has no valid columns.
TrainingDataNotCurrent: The training data have been changed since the model was trained; thus, the data are not imported as baseline data. To upload the correct baseline data, go to the Models tab in the deployment area or upload the data via the Model Repository.