Skip to main content
Version: Next

AI Conductor Terminology


Variable NameVariable DefinitionVariable Value
AWS_ACCOUNT_IDIdentifier for the AWS account (12-digit number)
AWS_ACCESS_KEY_IDACCESS KEY ID of the AWS account with admin privileges
AWS_SECRET_ACCESS_KEYSECRET ACCESS KEY of the AWS account with admin privileges
AWS_DEFAULT_REGIONAWS installation regione.g., ap-northeast-2 (Seoul)
AWS_DEFAULT_REGION_ALIASAbbreviation of the AWS installation regione.g., an2 (Seoul)
AWS_OUTPUT_FORMATAWS account output formatjson
DOCKER_NAMEName of the Docker Container for AWS Infra installation-
INFRA_NAMEName of the AWS Infra-
DEPLOY_ENVmellerikat development(dev)/production(prod) selectionChoose between dev or prod
AWS_CLUSTER_VERSIONAWS EKS versione.g., 1.28
AWS_CLUSTER_VERSION_NUMAWS EKS version number`echo ${AWS_CLUSTER_VERSION} | tr -d '.'`
e.g., 128
AWS_CLUSTER_VERSION_STRAWS EKS version string`echo ${AWS_CLUSTER_VERSION} | tr '.' '-'`
e.g., 1-28
AWS_CLUSTER_NAMEAWS EKS nameeks-an2-{INFRA_NAME}-{DEPLOY_ENV}-{AWS_CLUSTER_VERSION_STR}-eks-master
AWS_VPC_NAMEAWS VPC nameeksctl-eks-{AWS_DEFAULT_REGION_ALIAS}-{CLUSTER_NAME}-{DEPLOY_ENV}-
{AWS_CLUSTER_VERSION_STR}-eks-master-cluster/VPC
KUBEFLOW_USER_NAMEKubeflow account name for installation and operationuser-kubeflow-an2-{INFRA_NAME}-{DEPLOY_ENV}
KUBEFLOW_USER_POLICY_NAMEKubeflow account policy namepolicy-an2-{INFRA_NAME}-{DEPLOY_ENV}-s3-rwd
KUBEFLOW_USER_AWS_ACCESS_KEY_IDACCESS KEY ID of the user created for use in Kubeflow-
KUBEFLOW_USER_AWS_SECRET_ACCESS_KEYSECRET ACCESS KEY of the user created for use in Kubeflow-
EBS_CSI_SA_ROLE_NAMEAmazon EBS CSI driver role namerole-ebs-csidriver-{CLUSTER_NAME}
EBS_CSI_SA_ROLE_ARNARN of the installed Amazon EBS CSI driverarn:aws:iam::{AWS_ACCOUNT_ID}:role/{EBS_CSI_SA_ROLE_NAME}
EFS_CSI_SA_ROLE_NAMEAmazon EFS CSI driver role namerole-efs-csidriver-{CLUSTER_NAME}
EFS_CSI_SA_ROLE_ARNARN of the installed Amazon EFS CSI driverarn:aws:iam::{AWS_ACCOUNT_ID}:role/{EFS_CSI_SA_ROLE_NAME}
DB_HOSTEndpoint of the created database-
DB_PORTPort of the created database3310
DB_ADMIN_USERNAMEAdmin account of the created databaseadmin
DB_ADMIN_PASSWORDPassword of the admin account of the created database-
DB_APP_USERNAMEDatabase account for use in mellerikat (uppercase)`echo ${INFRA_NAME}_APP | tr [:lower:] [:upper:]`
DB_APP_PASSWORDPassword of the database account for use in mellerikat-
REDIS_HOSTRedis Primary Endpoint-
REDIS_PORTRedis Port6379
AWS_SECRETS_MANAGER_S3Name of the AWS Secrets Manager S3/parameter/{AWS_DEFAULT_REGION}/{INFRA_NAME}/{DEPLOY_ENV}/s3
AWS_SECRETS_MANAGER_RDSName of the AWS Secrets Manager RDS/parameter/{AWS_DEFAULT_REGION}/{INFRA_NAME}/{DEPLOY_ENV}/rds
KUBEFLOW_VERSIONKubeflow Versione.g., 1.8
KUBEFLOW_RELEASE_VERSIONKubeflow Release Versione.g., v1.8-branch
KUBEFLOW_VERSION_NUMKubeflow Version number`echo ${KUBEFLOW_VERSION} | tr -d '.'`
e.g., 18
KUBEFLOW_INSTALL_VERSIONKubeflow install version based on EKSkf{KUBEFLOW_VERSION_NUM}-eks{AWS_CLUSTER_VERSION_NUM}
e.g., kf18-eks128
KUBEFLOW_KATIB_IMAGEKubeflow Katib image URI{AWS_ACCOUNT_ID}.dkr.ecr.{AWS_DEFAULT_REGION}.amazonaws.com/
ecr-repo-{AWS_DEFAULT_REGION_ALIAS}-{INFRA_NAME}-{DEPLOY_ENV}/kubeflowkatib/katib-db-manager
KUBEFLOW_OIDC_IMAGEKubeflow OIDC image URI{AWS_ACCOUNT_ID}.dkr.ecr.{AWS_DEFAULT_REGION}.amazonaws.com/
ecr-repo-{AWS_DEFAULT_REGION_ALIAS}-{INFRA_NAME}-{DEPLOY_ENV}/oidc-authservice
KUBEFLOW_API_SERVER_IMAGEKubeflow API Server image URI{AWS_ACCOUNT_ID}.dkr.ecr.{AWS_DEFAULT_REGION}.amazonaws.com/
ecr-repo-{AWS_DEFAULT_REGION_ALIAS}-{INFRA_NAME}-{DEPLOY_ENV}/ml-pipeline/api-server
KUBEFLOW_CACHE_SERVER_IMAGEKubeflow Cache Server image URI{AWS_ACCOUNT_ID}.dkr.ecr.{AWS_DEFAULT_REGION}.amazonaws.com/
ecr-repo-{AWS_DEFAULT_REGION_ALIAS}-{INFRA_NAME}-{DEPLOY_ENV}/ml-pipeline/cache-server
KUBEFLOW_MANIFEST_GIT_URLKubeflow installation manifest git URLhttps://github.com/mellerikat/kubeflow-on-aws-manifests.git
PROJECT_NAMEProject name-
WORKSPACE_NAMEProject workspace name{PROJECT_NAME}-ws
WORKSPACE_IDThe ID of the created Workspaceuuid format
PROJECT_DASHBOARD_HASHthe value of Workspace hash in Monitoring DashboardAdd Monitoring Dashboard
e.g. : adz19cb14fkzka
PROJECT_KUBEFLOW_NAMESPACE_NAMEKubeflow Namespace name matching the workspaceaic-ns-{WORKSPACE_NAME}
PROJECT_KUBEFLOW_USER_NAMEAccount name for accessing the Kubeflow Namespace matching the workspaceaic-user-{WORKSPACE_NAME}
PROJECT_KUBEFLOW_USER_UNIQUE_IDUnique ID of the Kubeflow Namespace access account`date +"%G%m%e%H%M%S"`
PROJECT_KUBEFLOW_USER_PASSWDPassword for accessing the Kubeflow Namespace matching the workspaceGenerate hash with aic@com at bcrypt-generator
For special character recognition, use {PROJECT_KUBEFLOW_USER_PASSWD}=''
PROJECT_S3_BUCKET_NAMES3 Bucket name to be used in the projects3-{AWS_DEFAULT_REGION_ALIAS}-{INFRA_NAME}-{DEPLOY_ENV}-{PROJECT_NAME}
PROJECT_NODEGROUP_SPECLimit Nodegroup needed for Model training to 6 for user convenience
- low, standard, high include CPU/MEM EC2 specs
- low-gpu, standard-gpu, high-gpu include CPU/MEM/GPU EC2 specs		low, standard, high
low-gpu, standard-gpu, high-gpu
PROJECT_NODEGROUP_LABELNodegroup Label{PROJECT_NAME}-ws-{PROJECT_NODEGROUP_SPEC}
PROJECT_NODEGROUP_NAMENodegroup name to be used in the projectng-{AWS_DEFAULT_REGION_ALIAS}-aicond-{PROJECT_NAME}-ws-{PROJECT_NODEGROUP_SPEC}
PROJECT_NODEGROUP_DESIRED_SIZENodegroup desired sizeNodegroup desired size >= 0
PROJECT_NODEGROUP_MINNodegroup min size0
PROJECT_NODEGROUP_MAXNodegroup max size> PROJECT_NODEGROUP_MIN && > PROJECT_NODEGROUP_DESIRED_SIZE
PROJECT_NODEGROUP_EC2_NAMENodegroup EC2 Instance Sizee.g., m5.large
PROJECT_NODEGROUP_EC2_VCPUNodegroup EC2 Instance Size vCPU-
PROJECT_NODEGROUP_EC2_MEMNodegroup EC2 Instance Size Memory(GiB)-
PROJECT_NODEGROUP_EC2_GPUNodegroup EC2 Instance Size GPUs-
AIC_RELEASE_VERSIONAI Conductor release versione.g., 1.9.1
AIC_FRONTEND_URLAI Conductor Frontend URLaicond-{DOCKER_NAME}
AIC_BACKEND_URLAI Conductor Backend URLaicond-{DOCKER_NAME}/api/v1/docs
AIC_KUBEFLOW_URLAI Conductor Kubeflow URLaicond-kf-{DOCKER_NAME}
AIC_BACKEND_IMAGE_URLAI Conductor Backend Image URL{AWS_ACCOUNT_ID}.dkr.ecr.{AWS_DEFAULT_REGION}.amazonaws.com/
ecr-repo-{AWS_DEFAULT_REGION_ALIAS}-{INFRA_NAME}-{DEPLOY_ENV}/ai-advisor/ai-conductor/aic-webserver
AIC_BACKEND_IMAGE_TAGAI Conductor Backend Image TAG-
AIC_FRONTEND_IMAGE_URLAI Conductor Backend Image URL{AWS_ACCOUNT_ID}.dkr.ecr.{AWS_DEFAULT_REGION}.amazonaws.com/
ecr-repo-{AWS_DEFAULT_REGION_ALIAS}-{INFRA_NAME}-{DEPLOY_ENV}/ai-advisor/ai-conductor/aic-frontend
AIC_FRONTEND_IMAGE_TAGAI Conductor Frontend Image TAG-
AIC_ADMIN_USERAdmin account or ADMIN privileged account name for AI Conductoradmin
AIC_ADMIN_PASSWDAdmin account or ADMIN privileged account password for AI Conductor-
AIC_ADMIN_PASSWD_ENCRYPTEncrypted admin account or ADMIN privileged account password for AI ConductorEncrypt {AIC_ADMIN_PASSWD} at bcrypt-generator
For special character recognition, use {AIC_ADMIN_PASSWD_ENCRYPT}=''
AIC_ADMIN_COMPANYAdmin company for AI Conductor-
AIC_ADMIN_DEPARTMENTAdmin department for AI Conductor-
AIC_ECR_BASE_PATHECR base path for public use in mellerikat{AWS_ACCOUNT_ID}.dkr.ecr.{AWS_DEFAULT_REGION}.amazonaws.com/
ecr-repo-{AWS_DEFAULT_REGION_ALIAS}-{INFRA_NAME}-{DEPLOY_ENV}/
AIC_ACCOUNT_NAMEAccount ID for AI Conductor-
AIC_ACCOUNT_PASSWDAccount passworkd for AI Conductor-
AIC_ACCOUNT_TYPEAccount type for AI ConductorADMIN or DEVELOPER
AIC_ACCOUNT_ROLEAccount role for AI ConductorADMIN or MANAGER or MEMBER
AIC_MLOPS_ENGINEMLOps enginekubeflow or sagemaker
DASHBOARD_PREFIXURL of the monitoring toolaicond-mon.{DOMAIN_NAME}.com/d/