api_doc.md

API Reference

Packages

• data.fluid.io/v1alpha1

data.fluid.io/v1alpha1

Package v1alpha1 is the v1alpha1 version of the API.

Package v1alpha1 contains API Schema definitions for the data v1alpha1 API group

zz_unit_test_scheme.go with a "zz" prefix to ensure its init function is called after all the other init function in the pacakge

Resource Types

• AlluxioRuntime
• CacheRuntime
• CacheRuntimeClass
• DataBackup
• DataLoad
• DataMigrate
• DataProcess
• Dataset
• EFCRuntime
• GooseFSRuntime
• JindoRuntime
• JuiceFSRuntime
• ThinRuntime
• ThinRuntimeProfile
• VineyardRuntime

APIGatewayStatus

API Gateway

Appears in:

• RuntimeStatus

Field	Description	Default	Validation
endpoint string	Endpoint for accessing

AffinityPolicy

Underlying type: string

AffinityPolicy the strategy for the affinity between Data Operation Pods.

Appears in:

• AffinityStrategy

Field	Description
``
Require
Prefer

AffinityStrategy

Appears in:

• OperationRef

Field	Description	Default	Validation
dependOn ObjectRef	Specifies the dependent preceding operation in a workflow. If not set, use the operation referred to by RunAfter.		Optional: {}
policy AffinityPolicy	Policy one of: "", "Require", "Prefer"		Optional: {}
prefers Prefer array
requires Require array

AlluxioCompTemplateSpec

AlluxioCompTemplateSpec is a description of the Alluxio commponents

Appears in:

• AlluxioRuntimeSpec

Field	Description	Default	Validation
replicas integer	Replicas is the desired number of replicas of the given template. If unspecified, defaults to 1. replicas is the min replicas of dataset in the cluster		Minimum: 1 Optional: {}
jvmOptions string array	Options for JVM
properties object (keys:string, values:string)	Configurable properties for the Alluxio component. Refer to Alluxio Configuration Properties for more info		Optional: {}
ports object (keys:string, values:integer)	Ports used by Alluxio(e.g. rpc: 19998 for master)		Optional: {}
resources ResourceRequirements	Resources that will be requested by the Alluxio component. Resources are not allowed for ephemeral containers. Ephemeral containers use spare resources already allocated to the pod.		Optional: {}
env object (keys:string, values:string)	Environment variables that will be used by Alluxio component.
enabled boolean	Enabled or Disabled for the components. For now, only API Gateway is enabled or disabled.		Optional: {}
nodeSelector object (keys:string, values:string)	NodeSelector is a selector which must be true for the master to fit on a node		Optional: {}
networkMode NetworkMode	Whether to use hostnetwork or not		Enum: [HostNetwork ContainerNetwork] Optional: {}
volumeMounts VolumeMount array	VolumeMounts specifies the volumes listed in ".spec.volumes" to mount into the alluxio runtime component's filesystem.		Optional: {}
podMetadata PodMetadata	PodMetadata defines labels and annotations that will be propagated to Alluxio's pods		Optional: {}
imagePullSecrets LocalObjectReference array	ImagePullSecrets that will be used to pull images

AlluxioFuseSpec

AlluxioFuseSpec is a description of the Alluxio Fuse

Appears in:

• AlluxioRuntimeSpec

Field	Description	Default	Validation
image string	Image for Alluxio Fuse(e.g. alluxio/alluxio-fuse)
imageTag string	Image Tag for Alluxio Fuse(e.g. 2.3.0-SNAPSHOT)
imagePullPolicy string	One of the three policies: Always, IfNotPresent, Never
imagePullSecrets LocalObjectReference array	ImagePullSecrets that will be used to pull images
jvmOptions string array	Options for JVM
properties object (keys:string, values:string)	Configurable properties for Alluxio System. Refer to Alluxio Configuration Properties for more info
env object (keys:string, values:string)	Environment variables that will be used by Alluxio Fuse
resources ResourceRequirements	Resources that will be requested by Alluxio Fuse. Resources are not allowed for ephemeral containers. Ephemeral containers use spare resources already allocated to the pod.		Optional: {}
args string array	Arguments that will be passed to Alluxio Fuse
nodeSelector object (keys:string, values:string)	NodeSelector is a selector which must be true for the fuse client to fit on a node, this option only effect when global is enabled		Optional: {}
cleanPolicy FuseCleanPolicy	CleanPolicy decides when to clean Alluxio Fuse pods. Currently Fluid supports two policies: OnDemand and OnRuntimeDeleted OnDemand cleans fuse pod once the fuse pod on some node is not needed OnRuntimeDeleted cleans fuse pod only when the cache runtime is deleted Defaults to OnRuntimeDeleted		Optional: {}
networkMode NetworkMode	Whether to use hostnetwork or not		Enum: [HostNetwork ContainerNetwork] Optional: {}
volumeMounts VolumeMount array	VolumeMounts specifies the volumes listed in ".spec.volumes" to mount into the alluxio runtime component's filesystem.		Optional: {}
podMetadata PodMetadata	PodMetadata defines labels and annotations that will be propagated to Alluxio's fuse pods		Optional: {}

AlluxioRuntime

AlluxioRuntime is the Schema for the alluxioruntimes API

Field	Description	Default	Validation
apiVersion string	data.fluid.io/v1alpha1
kind string	AlluxioRuntime
metadata ObjectMeta	Refer to Kubernetes API documentation for fields of metadata.
spec AlluxioRuntimeSpec

AlluxioRuntimeSpec

AlluxioRuntimeSpec defines the desired state of AlluxioRuntime

Appears in:

• AlluxioRuntime

Field	Description	Default	Validation
alluxioVersion VersionSpec	The version information that instructs fluid to orchestrate a particular version of Alluxio.
master AlluxioCompTemplateSpec	The component spec of Alluxio master
jobMaster AlluxioCompTemplateSpec	The component spec of Alluxio job master
worker AlluxioCompTemplateSpec	The component spec of Alluxio worker
jobWorker AlluxioCompTemplateSpec	The component spec of Alluxio job Worker
apiGateway AlluxioCompTemplateSpec	The component spec of Alluxio API Gateway
initUsers InitUsersSpec	The spec of init users
fuse AlluxioFuseSpec	The component spec of Alluxio Fuse
properties object (keys:string, values:string)	Configurable properties for Alluxio system. Refer to Alluxio Configuration Properties for more info
jvmOptions string array	Options for JVM
tieredstore TieredStore	Tiered storage used by Alluxio
data Data	Management strategies for the dataset to which the runtime is bound
replicas integer	The replicas of the worker, need to be specified
runAs User	Manage the user to run Alluxio Runtime
disablePrometheus boolean	Disable monitoring for Alluxio Runtime Prometheus is enabled by default		Optional: {}
hadoopConfig string	Name of the configMap used to support HDFS configurations when using HDFS as Alluxio's UFS. The configMap must be in the same namespace with the AlluxioRuntime. The configMap should contain user-specific HDFS conf files in it. For now, only "hdfs-site.xml" and "core-site.xml" are supported. It must take the filename of the conf file as the key and content of the file as the value.		Optional: {}
volumes Volume array	Volumes is the list of Kubernetes volumes that can be mounted by the alluxio runtime components and/or fuses.		Optional: {}
podMetadata PodMetadata	PodMetadata defines labels and annotations that will be propagated to Alluxio's pods		Optional: {}
management RuntimeManagement	RuntimeManagement defines policies when managing the runtime		Optional: {}
imagePullSecrets LocalObjectReference array	ImagePullSecrets that will be used to pull images		Optional: {}

CacheRuntime

CacheRuntime is the Schema for the CacheRuntimes API

Field	Description	Default	Validation
apiVersion string	data.fluid.io/v1alpha1
kind string	CacheRuntime
metadata ObjectMeta	Refer to Kubernetes API documentation for fields of metadata.
spec CacheRuntimeSpec

CacheRuntimeClass

CacheRuntimeClass is the Schema for the cacheruntimeclasses API. CacheRuntimeClass defines a class of cache runtime implementations with specific configurations.

Field	Description	Default	Validation
apiVersion string	data.fluid.io/v1alpha1
kind string	CacheRuntimeClass
metadata ObjectMeta	Refer to Kubernetes API documentation for fields of metadata.
fileSystemType string	FileSystemType is the file system type of the cache runtime (e.g., "alluxio", "juicefs")		Required: {}
topology RuntimeTopology	Topology describes the topology of the CacheRuntime components (master, worker, client)		Optional: {}
extraResources RuntimeExtraResources	ExtraResources specifies additional resources (e.g., ConfigMaps) used by the CacheRuntime components		Optional: {}

CacheRuntimeClientSpec

CacheRuntimeClientSpec describes the desired state of CacheRuntime client component

Appears in:

• CacheRuntimeSpec

Field	Description	Default	Validation
disabled boolean	Disabled indicates whether the component should be disabled. If set to true, the component will not be created.		Optional: {}
runtimeVersion VersionSpec	RuntimeVersion is the version information that instructs Fluid to orchestrate a particular version of the runtime.		Optional: {}
options object (keys:string, values:string)	Options is a set of key-value pairs that provide additional configuration for the cache system.		Optional: {}
podMetadata PodMetadata	PodMetadata contains labels and annotations that will be propagated to the component's pods.		Optional: {}
resources ResourceRequirements	Resources describes the compute resource requirements. More info: https://kubernetes.io/docs/concepts/configuration/manage-resources-containers/		Optional: {}
env EnvVar array	Env is a list of environment variables to set in the container.		Optional: {}
volumeMounts VolumeMount array	VolumeMounts are the volumes to mount into the container's filesystem. Cannot be updated.		Optional: {}
args string array	Args are arguments to the entrypoint. The container image's CMD is used if this is not provided.		Optional: {}
nodeSelector object (keys:string, values:string)	NodeSelector is a selector which must be true for the pod to fit on a node. Selector which must match a node's labels for the pod to be scheduled on that node. More info: https://kubernetes.io/docs/concepts/configuration/assign-pod-node/		Optional: {}
tolerations Toleration array	Tolerations are the pod's tolerations. If specified, the pod's tolerations. More info: https://kubernetes.io/docs/concepts/scheduling-eviction/taint-and-toleration/		Optional: {}
tieredStore RuntimeTieredStore	TieredStore describes the tiered storage configuration used by the client component.		Optional: {}
cleanPolicy FuseCleanPolicy	CleanPolicy determines when to clean up the FUSE client pods. Currently supports two policies: - "OnDemand": Clean up the FUSE pod when it is no longer needed on a node - "OnRuntimeDeleted": Clean up the FUSE pod only when the CacheRuntime is deleted Defaults to "OnRuntimeDeleted".	OnRuntimeDeleted	Enum: [OnRuntimeDeleted OnDemand] Optional: {}

CacheRuntimeMasterSpec

CacheRuntimeMasterSpec describes the desired state of CacheRuntime master component

Appears in:

• CacheRuntimeSpec

Field	Description	Default	Validation
disabled boolean	Disabled indicates whether the component should be disabled. If set to true, the component will not be created.		Optional: {}
runtimeVersion VersionSpec	RuntimeVersion is the version information that instructs Fluid to orchestrate a particular version of the runtime.		Optional: {}
options object (keys:string, values:string)	Options is a set of key-value pairs that provide additional configuration for the cache system.		Optional: {}
podMetadata PodMetadata	PodMetadata contains labels and annotations that will be propagated to the component's pods.		Optional: {}
resources ResourceRequirements	Resources describes the compute resource requirements. More info: https://kubernetes.io/docs/concepts/configuration/manage-resources-containers/		Optional: {}
env EnvVar array	Env is a list of environment variables to set in the container.		Optional: {}
volumeMounts VolumeMount array	VolumeMounts are the volumes to mount into the container's filesystem. Cannot be updated.		Optional: {}
args string array	Args are arguments to the entrypoint. The container image's CMD is used if this is not provided.		Optional: {}
nodeSelector object (keys:string, values:string)	NodeSelector is a selector which must be true for the pod to fit on a node. Selector which must match a node's labels for the pod to be scheduled on that node. More info: https://kubernetes.io/docs/concepts/configuration/assign-pod-node/		Optional: {}
tolerations Toleration array	Tolerations are the pod's tolerations. If specified, the pod's tolerations. More info: https://kubernetes.io/docs/concepts/scheduling-eviction/taint-and-toleration/		Optional: {}
replicas integer	Replicas is the desired number of replicas of the master component. If unspecified, defaults to 1.	1	Minimum: 0 Optional: {}

CacheRuntimeSpec

CacheRuntimeSpec describes the desired state of CacheRuntime

Appears in:

• CacheRuntime

Field	Description	Default	Validation
runtimeClassName string	RuntimeClassName is the name of the CacheRuntimeClass to use for this runtime. The CacheRuntimeClass defines the implementation details of the cache runtime.		Required: {}
master CacheRuntimeMasterSpec	Master is the desired state of the master component.		Optional: {}
worker CacheRuntimeWorkerSpec	Worker is the desired state of the worker component.		Optional: {}
client CacheRuntimeClientSpec	Client is the desired state of the client component.		Optional: {}
options object (keys:string, values:string)	Options is a set of key-value pairs that provide additional configuration for the cache system. These options will be propagated to all components.		Optional: {}
podMetadata PodMetadata	PodMetadata contains labels and annotations that will be propagated to all component pods.		Optional: {}
imagePullSecrets LocalObjectReference array	ImagePullSecrets is an optional list of references to secrets in the same namespace to use for pulling any of the images used by this PodSpec. More info: https://kubernetes.io/docs/concepts/containers/images#specifying-imagepullsecrets-on-a-pod		Optional: {}
volumes Volume array	Volumes is the list of volumes that can be mounted by containers belonging to the cache runtime components. More info: https://kubernetes.io/docs/concepts/storage/volumes		Optional: {}

CacheRuntimeWorkerSpec

CacheRuntimeWorkerSpec describes the desired state of CacheRuntime worker component

Appears in:

• CacheRuntimeSpec

Field	Description	Default	Validation
disabled boolean	Disabled indicates whether the component should be disabled. If set to true, the component will not be created.		Optional: {}
runtimeVersion VersionSpec	RuntimeVersion is the version information that instructs Fluid to orchestrate a particular version of the runtime.		Optional: {}
options object (keys:string, values:string)	Options is a set of key-value pairs that provide additional configuration for the cache system.		Optional: {}
podMetadata PodMetadata	PodMetadata contains labels and annotations that will be propagated to the component's pods.		Optional: {}
resources ResourceRequirements	Resources describes the compute resource requirements. More info: https://kubernetes.io/docs/concepts/configuration/manage-resources-containers/		Optional: {}
env EnvVar array	Env is a list of environment variables to set in the container.		Optional: {}
volumeMounts VolumeMount array	VolumeMounts are the volumes to mount into the container's filesystem. Cannot be updated.		Optional: {}
args string array	Args are arguments to the entrypoint. The container image's CMD is used if this is not provided.		Optional: {}
nodeSelector object (keys:string, values:string)	NodeSelector is a selector which must be true for the pod to fit on a node. Selector which must match a node's labels for the pod to be scheduled on that node. More info: https://kubernetes.io/docs/concepts/configuration/assign-pod-node/		Optional: {}
tolerations Toleration array	Tolerations are the pod's tolerations. If specified, the pod's tolerations. More info: https://kubernetes.io/docs/concepts/scheduling-eviction/taint-and-toleration/		Optional: {}
replicas integer	Replicas is the desired number of replicas of the worker component. If unspecified, defaults to 1.	1	Minimum: 0 Optional: {}
tieredStore RuntimeTieredStore	TieredStore describes the tiered storage configuration used by the worker component.		Optional: {}

CacheableNodeAffinity

CacheableNodeAffinity defines constraints that limit what nodes this dataset can be cached to.

Appears in:

• DatasetSpec

Field	Description	Default	Validation
required NodeSelector	Required specifies hard node constraints that must be met.

CleanCachePolicy

CleanCachePolicy defines policies when cleaning cache

Appears in:

• EFCRuntimeSpec
• GooseFSRuntimeSpec
• JindoRuntimeSpec
• RuntimeManagement

Field	Description	Default	Validation
gracePeriodSeconds integer	Optional duration in seconds the cache needs to clean gracefully. May be decreased in delete runtime request. Value must be non-negative integer. The value zero indicates clean immediately via the timeout command (no opportunity to shut down). If this value is nil, the default grace period will be used instead. The grace period is the duration in seconds after the processes running in the pod are sent a termination signal and the time when the processes are forcibly halted with timeout command. Set this value longer than the expected cleanup time for your process.	60	Optional: {}
maxRetryAttempts integer	Optional max retry Attempts when cleanCache function returns an error after execution, runtime attempts to run it three more times by default. With Maximum Retry Attempts, you can customize the maximum number of retries. This gives you the option to continue processing retries.	3	Optional: {}

ClientMetrics

Appears in:

• JindoFuseSpec

Field	Description	Default	Validation
enabled boolean	Enabled decides whether to expose client metrics.
scrapeTarget string	ScrapeTarget decides which fuse component will be scraped by Prometheus. It is a list separated by comma where supported items are [MountPod, Sidecar, All (indicates MountPod and Sidecar), None]. Defaults to None when it is not explicitly set.

ComponentServiceConfig

ComponentServiceConfig defines the service configuration for runtime components. Currently only headless service is supported, but this can be extended in the future to support other service types such as ClusterIP, NodePort, LoadBalancer, etc.

Appears in:

• RuntimeComponentService

Field	Description	Default	Validation
headless HeadlessRuntimeComponentService	Headless enables a headless service for the component. A headless service does not allocate a cluster IP and allows direct pod-to-pod communication.		Optional: {}

Condition

Condition explains the transitions on phase

Appears in:

• OperationStatus

Field	Description	Default	Validation
type ConditionType	Type of condition, either Complete or Failed
reason string	Reason for the condition's last transition
message string	Message is a human-readable message indicating details about the transition
lastProbeTime Time	LastProbeTime describes last time this condition was updated.
lastTransitionTime Time	LastTransitionTime describes last time the condition transitioned from one status to another.

ConfigMapDependencyConfig

ConfigMapDependencyConfig defines the ConfigMap mount configuration

Appears in:

• ExtraResourcesComponentDependency

Field	Description	Default	Validation
name string	Name is the ConfigMap template name defined in extraResources.configMaps		Optional: {}
mountPath string	MountPath is the path within the container at which the ConfigMap should be mounted. Must not contain ':'.		Optional: {}

ConfigMapRuntimeExtraResource

ConfigMapRuntimeExtraResource defines a ConfigMap template for CacheRuntime extra resources

Appears in:

• RuntimeExtraResources

Field	Description	Default	Validation
name string	Name is the name of the ConfigMap. This will be used as the actual ConfigMap name when created in the runtime's namespace.		Optional: {}
data object (keys:string, values:string)	Data contains the configuration data. Each key must consist of alphanumeric characters, '-', '_' or '.'. Values with non-UTF-8 byte sequences must use the BinaryData field. The keys stored in Data must not overlap with the keys in the BinaryData field, this is enforced during validation process.		Optional: {}

Data

Data management strategies

Appears in:

• AlluxioRuntimeSpec
• GooseFSRuntimeSpec

Field	Description	Default	Validation
replicas integer	The copies of the dataset		Optional: {}
pin boolean	Pin the dataset or not. Refer to Alluxio User-CLI pin		Optional: {}

DataBackup

DataBackup is the Schema for the backup API

Field	Description	Default	Validation
apiVersion string	data.fluid.io/v1alpha1
kind string	DataBackup
metadata ObjectMeta	Refer to Kubernetes API documentation for fields of metadata.
spec DataBackupSpec

DataBackupSpec

DataBackupSpec defines the desired state of DataBackup

Appears in:

• DataBackup

Field	Description	Default	Validation
dataset string	Dataset defines the target dataset of the DataBackup
backupPath string	BackupPath defines the target path to save data of the DataBackup
runAs User	Manage the user to run Alluxio DataBackup
runAfter OperationRef	Specifies that the preceding operation in a workflow		Optional: {}
ttlSecondsAfterFinished integer	TTLSecondsAfterFinished is the time second to clean up data operations after finished or failed		Optional: {}

DataLoad

DataLoad is the Schema for the dataloads API

Field	Description	Default	Validation
apiVersion string	data.fluid.io/v1alpha1
kind string	DataLoad
metadata ObjectMeta	Refer to Kubernetes API documentation for fields of metadata.
spec DataLoadSpec

DataLoadSpec

DataLoadSpec defines the desired state of DataLoad

Appears in:

• DataLoad

Field	Description	Default	Validation
dataset TargetDataset	Dataset defines the target dataset of the DataLoad
loadMetadata boolean	LoadMetadata specifies if the dataload job should load metadata
target TargetPath array	Target defines target paths that needs to be loaded
options object (keys:string, values:string)	Options specifies the extra dataload properties for runtime
podMetadata PodMetadata	PodMetadata defines labels and annotations that will be propagated to DataLoad pods
affinity Affinity	Affinity defines affinity for DataLoad pod		Optional: {}
tolerations Toleration array	Tolerations defines tolerations for DataLoad pod		Optional: {}
nodeSelector object (keys:string, values:string)	NodeSelector defiens node selector for DataLoad pod		Optional: {}
schedulerName string	SchedulerName sets the scheduler to be used for DataLoad pod		Optional: {}
policy Policy	including Once, Cron, OnEvent	Once	Enum: [Once Cron OnEvent] Optional: {}
schedule string	The schedule in Cron format, only set when policy is cron, see https://en.wikipedia.org/wiki/Cron.		Optional: {}
runAfter OperationRef	Specifies that the preceding operation in a workflow		Optional: {}
ttlSecondsAfterFinished integer	TTLSecondsAfterFinished is the time second to clean up data operations after finished or failed		Optional: {}
resources ResourceRequirements	Resources that will be requested by the DataLoad job.		Optional: {}

DataMigrate

DataMigrate is the Schema for the datamigrates API

Field	Description	Default	Validation
apiVersion string	data.fluid.io/v1alpha1
kind string	DataMigrate
metadata ObjectMeta	Refer to Kubernetes API documentation for fields of metadata.
spec DataMigrateSpec

DataMigrateSpec

DataMigrateSpec defines the desired state of DataMigrate

Appears in:

• DataMigrate

Field	Description	Default	Validation
image string	Image (e.g. alluxio/alluxio)
imageTag string	Image tag (e.g. 2.3.0-SNAPSHOT)
imagePullPolicy string	One of the three policies: Always, IfNotPresent, Never
from DataToMigrate	data to migrate source, including dataset and external storage
to DataToMigrate	data to migrate destination, including dataset and external storage
block boolean	if dataMigrate blocked dataset usage, default is false		Optional: {}
runtimeType string	using which runtime to migrate data; if none, take dataset runtime as default		Optional: {}
options object (keys:string, values:string)	options for migrate, different for each runtime		Optional: {}
policy Policy	policy for migrate, including Once, Cron, OnEvent	Once	Enum: [Once Cron OnEvent] Optional: {}
schedule string	The schedule in Cron format, only set when policy is cron, see https://en.wikipedia.org/wiki/Cron.		Optional: {}
podMetadata PodMetadata	PodMetadata defines labels and annotations that will be propagated to DataMigrate pods
affinity Affinity	Affinity defines affinity for DataMigrate pod		Optional: {}
tolerations Toleration array	Tolerations defines tolerations for DataMigrate pod		Optional: {}
nodeSelector object (keys:string, values:string)	NodeSelector defiens node selector for DataMigrate pod		Optional: {}
schedulerName string	SchedulerName sets the scheduler to be used for DataMigrate pod		Optional: {}
runAfter OperationRef	Specifies that the preceding operation in a workflow		Optional: {}
ttlSecondsAfterFinished integer	TTLSecondsAfterFinished is the time second to clean up data operations after finished or failed		Optional: {}
resources ResourceRequirements	Resources that will be requested by the DataMigrate job.		Optional: {}
parallelism integer	Parallelism defines the parallelism tasks numbers for DataMigrate. If the value is greater than 1, the job acts as a launcher, and users should define the WorkerSpec.	1	Minimum: 1 Optional: {}
parallelOptions object (keys:string, values:string)	ParallelOptions defines options like ssh port and ssh secret name when parallelism is greater than 1.		Optional: {}

DataProcess

DataProcess is the Schema for the dataprocesses API

Field	Description	Default	Validation
apiVersion string	data.fluid.io/v1alpha1
kind string	DataProcess
metadata ObjectMeta	Refer to Kubernetes API documentation for fields of metadata.
spec DataProcessSpec

DataProcessSpec

DataProcessSpec defines the desired state of DataProcess

Appears in:

• DataProcess

Field	Description	Default	Validation
dataset TargetDatasetWithMountPath	Dataset specifies the target dataset and its mount path.		Required: {}
processor Processor	Processor specify how to process data.		Required: {}
runAfter OperationRef	Specifies that the preceding operation in a workflow		Optional: {}
ttlSecondsAfterFinished integer	TTLSecondsAfterFinished is the time second to clean up data operations after finished or failed		Optional: {}

DataRestoreLocation

DataRestoreLocation describes the spec restore location of Dataset

Appears in:

• DatasetSpec

Field	Description	Default	Validation
path string	Path describes the path of restore, in the form of local://subpath or pvc:///subpath		Optional: {}
nodeName string	NodeName describes the nodeName of restore if Path is in the form of local://subpath		Optional: {}

DataToMigrate

Appears in:

• DataMigrateSpec

Field	Description	Default	Validation
dataset DatasetToMigrate	dataset to migrate
externalStorage ExternalStorage	external storage for data migrate

Dataset

Dataset is the Schema for the datasets API

Field	Description	Default	Validation
apiVersion string	data.fluid.io/v1alpha1
kind string	Dataset
metadata ObjectMeta	Refer to Kubernetes API documentation for fields of metadata.
spec DatasetSpec

DatasetCondition

Condition describes the state of the cache at a certain point.

Appears in:

• DatasetStatus

Field	Description	Default	Validation
type DatasetConditionType	Type of cache condition.
reason string	The reason for the condition's last transition.
message string	A human readable message indicating details about the transition.
lastUpdateTime Time	The last time this condition was updated.
lastTransitionTime Time	Last time the condition transitioned from one status to another.

DatasetConditionType

Underlying type: string

DatasetConditionType defines all kinds of types of cacheStatus.
one of the three types: RuntimeScheduled, Ready and Initialized

Appears in:

• DatasetCondition

Field	Description
RuntimeScheduled	RuntimeScheduled means the runtime CRD has been accepted by the system, But master and workers are not ready
Ready	DatasetReady means the cache system for the dataset is ready.
NotReady	DatasetNotReady means the dataset is not bound due to some unexpected error
UpdateReady	DatasetUpdateReady means the cache system for the dataset is updated.
Updating	DatasetUpdating means the cache system for the dataset is updating.
Initialized	DatasetInitialized means the cache system for the dataset is Initialized.

DatasetPhase

Underlying type: string

DatasetPhase indicates whether the loading is behaving

Appears in:

• DatasetStatus

Field	Description
Pending	TODO: add the Pending phase to Dataset
Bound	Bound to dataset, can't be released
Failed	Failed, can't be deleted
NotBound	Not bound to runtime, can be deleted
Updating	updating dataset, can't be released
DataMigrating	migrating dataset, can't be mounted
``	the dataset have no phase and need to be judged

DatasetSpec

DatasetSpec defines the desired state of Dataset

Appears in:

• Dataset

Field	Description	Default	Validation
mounts Mount array	Mount Points to be mounted on cache runtime. This field can be empty because some runtimes don't need to mount external storage (e.g. Vineyard).		MinItems: 1 UniqueItems: false Optional: {}
owner User	The owner of the dataset		Optional: {}
nodeAffinity CacheableNodeAffinity	NodeAffinity defines constraints that limit what nodes this dataset can be cached to. This field influences the scheduling of pods that use the cached dataset.		Optional: {}
tolerations Toleration array	If specified, the pod's tolerations.		Optional: {}
accessModes PersistentVolumeAccessMode array	AccessModes contains all ways the volume backing the PVC can be mounted		Optional: {}
runtimes Runtime array	Runtimes for supporting dataset (e.g. AlluxioRuntime)
placement PlacementMode	Manage switch for opening Multiple datasets single node deployment or not		Enum: [Exclusive Shared] Optional: {}
dataRestoreLocation DataRestoreLocation	DataRestoreLocation is the location to load data of dataset been backuped		Optional: {}
sharedOptions object (keys:string, values:string)	SharedOptions is the options to all mount		Optional: {}
sharedEncryptOptions EncryptOption array	SharedEncryptOptions is the encryptOption to all mount		Optional: {}

DatasetToMigrate

Appears in:

• DataToMigrate

Field	Description	Default	Validation
name string	name of dataset
namespace string	namespace of dataset
path string	path to migrate

EFCCompTemplateSpec

EFCCompTemplateSpec is a description of the EFC components

Appears in:

• EFCRuntimeSpec

Field	Description	Default	Validation
replicas integer	Replicas is the desired number of replicas of the given template. If unspecified, defaults to 1. replicas is the min replicas of dataset in the cluster		Minimum: 1 Optional: {}
version VersionSpec	The version information that instructs fluid to orchestrate a particular version of EFC Comp
properties object (keys:string, values:string)	Configurable properties for the EFC component.		Optional: {}
ports object (keys:string, values:integer)	Ports used by EFC(e.g. rpc: 19998 for master).		Optional: {}
resources ResourceRequirements	Resources that will be requested by the EFC component. Resources are not allowed for ephemeral containers. Ephemeral containers use spare resources already allocated to the pod.		Optional: {}
disabled boolean	Enabled or Disabled for the components. Default enable.		Optional: {}
nodeSelector object (keys:string, values:string)	NodeSelector is a selector which must be true for the component to fit on a node.		Optional: {}
networkMode NetworkMode	Whether to use host network or not.		Enum: [HostNetwork ContainerNetwork] Optional: {}
podMetadata PodMetadata	PodMetadata defines labels and annotations that will be propagated to EFC's master and worker pods		Optional: {}

EFCFuseSpec

EFCFuseSpec is a description of the EFC Fuse

Appears in:

• EFCRuntimeSpec

Field	Description	Default	Validation
version VersionSpec	The version information that instructs fluid to orchestrate a particular version of EFC Fuse
properties object (keys:string, values:string)	Configurable properties for EFC fuse		Optional: {}
resources ResourceRequirements	Resources that will be requested by EFC Fuse. Resources are not allowed for ephemeral containers. Ephemeral containers use spare resources already allocated to the pod.		Optional: {}
nodeSelector object (keys:string, values:string)	NodeSelector is a selector which must be true for the fuse client to fit on a node, this option only effect when global is enabled		Optional: {}
cleanPolicy FuseCleanPolicy	CleanPolicy decides when to clean EFC Fuse pods. Currently Fluid supports two policies: OnDemand and OnRuntimeDeleted OnDemand cleans fuse pod once th fuse pod on some node is not needed OnRuntimeDeleted cleans fuse pod only when the cache runtime is deleted Defaults to OnRuntimeDeleted		Optional: {}
networkMode NetworkMode	Whether to use hostnetwork or not		Enum: [HostNetwork ContainerNetwork] Optional: {}
podMetadata PodMetadata	PodMetadata defines labels and annotations that will be propagated to EFC's fuse pods		Optional: {}

EFCRuntime

EFCRuntime is the Schema for the efcruntimes API

Field	Description	Default	Validation
apiVersion string	data.fluid.io/v1alpha1
kind string	EFCRuntime
metadata ObjectMeta	Refer to Kubernetes API documentation for fields of metadata.
spec EFCRuntimeSpec

EFCRuntimeSpec

EFCRuntimeSpec defines the desired state of EFCRuntime

Appears in:

• EFCRuntime

Field	Description	Default	Validation
master EFCCompTemplateSpec	The component spec of EFC master
worker EFCCompTemplateSpec	The component spec of EFC worker
initFuse InitFuseSpec	The spec of init alifuse
fuse EFCFuseSpec	The component spec of EFC Fuse
tieredstore TieredStore	Tiered storage used by EFC worker
replicas integer	The replicas of the worker, need to be specified
osAdvise OSAdvise	Operating system optimization for EFC
cleanCachePolicy CleanCachePolicy	CleanCachePolicy defines cleanCache Policy		Optional: {}
podMetadata PodMetadata	PodMetadata defines labels and annotations that will be propagated to all EFC's pods		Optional: {}

EncryptOption

Appears in:

• DatasetSpec
• ExternalEndpointSpec
• ExternalStorage
• Mount

Field	Description	Default	Validation
name string	The name of encryptOption		Required: {}
valueFrom EncryptOptionSource	The valueFrom of encryptOption		Optional: {}

EncryptOptionComponentDependency

EncryptOptionComponentDependency defines the configuration for encrypt option dependency

Appears in:

• RuntimeComponentDependencies

EncryptOptionSource

Appears in:

• EncryptOption

Field	Description	Default	Validation
secretKeyRef SecretKeySelector	The encryptInfo obtained from secret		Optional: {}

ExternalEndpointSpec

ExternalEndpointSpec defines the configurations for external etcd cluster

Appears in:

• MasterSpec

Field	Description	Default	Validation
uri string	URI specifies the endpoint of external Etcd cluster E,g. "etcd-svc.etcd-namespace.svc.cluster.local:2379" Default is not set and use http protocol to connect to external etcd cluster		Optional: {}
encryptOptions EncryptOption array	encrypt info for accessing the external etcd cluster		Optional: {}
options object (keys:string, values:string)	Configurable options for External Etcd cluster.		Optional: {}

ExternalStorage

Appears in:

• DataToMigrate

Field	Description	Default	Validation
uri string	type of external storage, including s3, oss, gcs, ceph, nfs, pvc, etc. (related to runtime)
encryptOptions EncryptOption array	encrypt info for external storage		Optional: {}

ExtraResourcesComponentDependency

ExtraResourcesComponentDependency defines the extra resources configuration for component dependencies

Appears in:

• RuntimeComponentDependencies

Field	Description	Default	Validation
configMaps ConfigMapDependencyConfig array	ConfigMaps is a list of ConfigMaps in the same namespace to mount into the component		Optional: {}

FuseCleanPolicy

Underlying type: string

Appears in:

• AlluxioFuseSpec
• CacheRuntimeClientSpec
• EFCFuseSpec
• GooseFSFuseSpec
• JindoFuseSpec
• JuiceFSFuseSpec
• ThinFuseSpec
• VineyardClientSocketSpec

Field	Description
``	NoneCleanPolicy is the default clean policy. It will be transformed to OnRuntimeDeletedCleanPolicy automatically.
OnDemand	OnDemandCleanPolicy cleans fuse pod once the fuse pod on some node is not needed
OnRuntimeDeleted	OnRuntimeDeletedCleanPolicy cleans fuse pod only when the cache runtime is deleted
OnFuseChanged	OnFuseChangedCleanPolicy cleans fuse pod when the fuse in runtime is updated and the fuse pod on some node is not needed

GooseFSCompTemplateSpec

GooseFSCompTemplateSpec is a description of the GooseFS commponents

Appears in:

• GooseFSRuntimeSpec

Field	Description	Default	Validation
replicas integer	Replicas is the desired number of replicas of the given template. If unspecified, defaults to 1. replicas is the min replicas of dataset in the cluster		Minimum: 1 Optional: {}
jvmOptions string array	Options for JVM
properties object (keys:string, values:string)	Configurable properties for the GOOSEFS component. Refer to GOOSEFS Configuration Properties for more info		Optional: {}
ports object (keys:string, values:integer)	Ports used by GooseFS(e.g. rpc: 19998 for master)		Optional: {}
resources ResourceRequirements	Resources that will be requested by the GooseFS component. Resources are not allowed for ephemeral containers. Ephemeral containers use spare resources already allocated to the pod.		Optional: {}
env object (keys:string, values:string)	Environment variables that will be used by GooseFS component.
enabled boolean	Enabled or Disabled for the components. For now, only API Gateway is enabled or disabled.		Optional: {}
nodeSelector object (keys:string, values:string)	NodeSelector is a selector which must be true for the master to fit on a node		Optional: {}
annotations object (keys:string, values:string)	Annotations is an unstructured key value map stored with a resource that may be set by external tools to store and retrieve arbitrary metadata. They are not queryable and should be preserved when modifying objects. More info: http://kubernetes.io/docs/user-guide/annotations		Optional: {}

GooseFSFuseSpec

GooseFSFuseSpec is a description of the GooseFS Fuse

Appears in:

• GooseFSRuntimeSpec

Field	Description	Default	Validation
image string	Image for GooseFS Fuse(e.g. goosefs/goosefs-fuse)
imageTag string	Image Tag for GooseFS Fuse(e.g. v1.0.1)
imagePullPolicy string	One of the three policies: Always, IfNotPresent, Never
jvmOptions string array	Options for JVM
properties object (keys:string, values:string)	Configurable properties for the GOOSEFS component. Refer to GOOSEFS Configuration Properties for more info
env object (keys:string, values:string)	Environment variables that will be used by GooseFS Fuse
resources ResourceRequirements	Resources that will be requested by GooseFS Fuse. Resources are not allowed for ephemeral containers. Ephemeral containers use spare resources already allocated to the pod.		Optional: {}
args string array	Arguments that will be passed to GooseFS Fuse
nodeSelector object (keys:string, values:string)	NodeSelector is a selector which must be true for the fuse client to fit on a node, this option only effect when global is enabled		Optional: {}
cleanPolicy FuseCleanPolicy	CleanPolicy decides when to clean GooseFS Fuse pods. Currently Fluid supports two policies: OnDemand and OnRuntimeDeleted OnDemand cleans fuse pod once th fuse pod on some node is not needed OnRuntimeDeleted cleans fuse pod only when the cache runtime is deleted Defaults to OnRuntimeDeleted		Optional: {}
annotations object (keys:string, values:string)	Annotations is an unstructured key value map stored with a resource that may be set by external tools to store and retrieve arbitrary metadata. They are not queryable and should be preserved when modifying objects. More info: http://kubernetes.io/docs/user-guide/annotations		Optional: {}

GooseFSRuntime

GooseFSRuntime is the Schema for the goosefsruntimes API

Field	Description	Default	Validation
apiVersion string	data.fluid.io/v1alpha1
kind string	GooseFSRuntime
metadata ObjectMeta	Refer to Kubernetes API documentation for fields of metadata.
spec GooseFSRuntimeSpec

GooseFSRuntimeSpec

GooseFSRuntimeSpec defines the desired state of GooseFSRuntime

Appears in:

• GooseFSRuntime

Field	Description	Default	Validation
goosefsVersion VersionSpec	The version information that instructs fluid to orchestrate a particular version of GooseFS.
master GooseFSCompTemplateSpec	The component spec of GooseFS master
jobMaster GooseFSCompTemplateSpec	The component spec of GooseFS job master
worker GooseFSCompTemplateSpec	The component spec of GooseFS worker
jobWorker GooseFSCompTemplateSpec	The component spec of GooseFS job Worker
apiGateway GooseFSCompTemplateSpec	The component spec of GooseFS API Gateway
initUsers InitUsersSpec	The spec of init users
fuse GooseFSFuseSpec	The component spec of GooseFS Fuse
properties object (keys:string, values:string)	Configurable properties for the GOOSEFS component. Refer to GOOSEFS Configuration Properties for more info
jvmOptions string array	Options for JVM
tieredstore TieredStore	Tiered storage used by GooseFS
data Data	Management strategies for the dataset to which the runtime is bound
replicas integer	The replicas of the worker, need to be specified
runAs User	Manage the user to run GooseFS Runtime GooseFS support POSIX-ACL and Apache Ranger to manager authorization
disablePrometheus boolean	Disable monitoring for GooseFS Runtime Prometheus is enabled by default		Optional: {}
hadoopConfig string	Name of the configMap used to support HDFS configurations when using HDFS as GooseFS's UFS. The configMap must be in the same namespace with the GooseFSRuntime. The configMap should contain user-specific HDFS conf files in it. For now, only "hdfs-site.xml" and "core-site.xml" are supported. It must take the filename of the conf file as the key and content of the file as the value.		Optional: {}
cleanCachePolicy CleanCachePolicy	CleanCachePolicy defines cleanCache Policy		Optional: {}

HCFSStatus

HCFS Endpoint info

Appears in:

• DatasetStatus

Field	Description	Default	Validation
endpoint string	Endpoint for accessing
underlayerFileSystemVersion string	Underlayer HCFS Compatible Version

HeadlessRuntimeComponentService

HeadlessRuntimeComponentService defines the configuration for headless service

Appears in:

• ComponentServiceConfig
• RuntimeComponentService

InitFuseSpec

InitFuseSpec is a description of initialize the fuse kernel module for runtime

Appears in:

• EFCRuntimeSpec

Field	Description	Default	Validation
version VersionSpec	The version information that instructs fluid to orchestrate a particular version of Alifuse

InitUsersSpec

InitUsersSpec is a description of the initialize the users for runtime

Appears in:

• AlluxioRuntimeSpec
• GooseFSRuntimeSpec
• JuiceFSRuntimeSpec

Field	Description	Default	Validation
image string	Image for initialize the users for runtime(e.g. alluxio/alluxio-User init)
imageTag string	Image Tag for initialize the users for runtime(e.g. 2.3.0-SNAPSHOT)
imagePullPolicy string	One of the three policies: Always, IfNotPresent, Never
env object (keys:string, values:string)	Environment variables that will be used by initialize the users for runtime
resources ResourceRequirements	Resources that will be requested by initialize the users for runtime. Resources are not allowed for ephemeral containers. Ephemeral containers use spare resources already allocated to the pod.		Optional: {}

JindoCompTemplateSpec

JindoCompTemplateSpec is a description of the Jindo commponents

Appears in:

• JindoRuntimeSpec

Field	Description	Default	Validation
replicas integer	Replicas is the desired number of replicas of the given template. If unspecified, defaults to 1. replicas is the min replicas of dataset in the cluster		Minimum: 1 Optional: {}
properties object (keys:string, values:string)	Configurable properties for the Jindo component.		Optional: {}
ports object (keys:string, values:integer)			Optional: {}
resources ResourceRequirements	Resources that will be requested by the Jindo component. Resources are not allowed for ephemeral containers. Ephemeral containers use spare resources already allocated to the pod.		Optional: {}
env object (keys:string, values:string)	Environment variables that will be used by Jindo component.
nodeSelector object (keys:string, values:string)	NodeSelector is a selector which must be true for the master to fit on a node		Optional: {}
tolerations Toleration array	If specified, the pod's tolerations.		Optional: {}
labels object (keys:string, values:string)	Labels will be added on JindoFS Master or Worker pods. DEPRECATED: This is a deprecated field. Please use PodMetadata instead. Note: this field is set to be exclusive with PodMetadata.Labels		Optional: {}
podMetadata PodMetadata	PodMetadata defines labels and annotations that will be propagated to Jindo's pods		Optional: {}
disabled boolean	If disable JindoFS master or worker		Optional: {}
volumeMounts VolumeMount array	VolumeMounts specifies the volumes listed in ".spec.volumes" to mount into the jindo runtime component's filesystem.		Optional: {}
imagePullSecrets LocalObjectReference array	ImagePullSecrets that will be used to pull images

JindoFuseSpec

JindoFuseSpec is a description of the Jindo Fuse

Appears in:

• JindoRuntimeSpec

Field	Description	Default	Validation
image string	Image for Jindo Fuse(e.g. jindo/jindo-fuse)
imageTag string	Image Tag for Jindo Fuse(e.g. 2.3.0-SNAPSHOT)
imagePullPolicy string	One of the three policies: Always, IfNotPresent, Never
imagePullSecrets LocalObjectReference array	ImagePullSecrets that will be used to pull images
properties object (keys:string, values:string)	Configurable properties for Jindo System.
env object (keys:string, values:string)	Environment variables that will be used by Jindo Fuse
resources ResourceRequirements	Resources that will be requested by Jindo Fuse. Resources are not allowed for ephemeral containers. Ephemeral containers use spare resources already allocated to the pod.		Optional: {}
args string array	Arguments that will be passed to Jindo Fuse
nodeSelector object (keys:string, values:string)	NodeSelector is a selector which must be true for the fuse client to fit on a node, this option only effect when global is enabled		Optional: {}
tolerations Toleration array	If specified, the pod's tolerations.		Optional: {}
labels object (keys:string, values:string)	Labels will be added on all the JindoFS pods. DEPRECATED: this is a deprecated field. Please use PodMetadata.Labels instead. Note: this field is set to be exclusive with PodMetadata.Labels		Optional: {}
podMetadata PodMetadata	PodMetadata defines labels and annotations that will be propagated to Jindo's fuse pods		Optional: {}
cleanPolicy FuseCleanPolicy	CleanPolicy decides when to clean JindoFS Fuse pods. Currently Fluid supports two policies: OnDemand and OnRuntimeDeleted OnDemand cleans fuse pod once th fuse pod on some node is not needed OnRuntimeDeleted cleans fuse pod only when the cache runtime is deleted Defaults to OnRuntimeDeleted		Optional: {}
disabled boolean	If disable JindoFS fuse		Optional: {}
logConfig object (keys:string, values:string)			Optional: {}
metrics ClientMetrics	Define whether fuse metrics will be enabled.		Optional: {}

JindoRuntime

JindoRuntime is the Schema for the jindoruntimes API

Field	Description	Default	Validation
apiVersion string	data.fluid.io/v1alpha1
kind string	JindoRuntime
metadata ObjectMeta	Refer to Kubernetes API documentation for fields of metadata.
spec JindoRuntimeSpec

JindoRuntimeSpec

JindoRuntimeSpec defines the desired state of JindoRuntime

Appears in:

• JindoRuntime

Field	Description	Default	Validation
jindoVersion VersionSpec	The version information that instructs fluid to orchestrate a particular version of Jindo.
master JindoCompTemplateSpec	The component spec of Jindo master
worker JindoCompTemplateSpec	The component spec of Jindo worker
fuse JindoFuseSpec	The component spec of Jindo Fuse
properties object (keys:string, values:string)	Configurable properties for Jindo system.
tieredstore TieredStore	Tiered storage used by Jindo
replicas integer	The replicas of the worker, need to be specified
runAs User	Manage the user to run Jindo Runtime
user string
hadoopConfig string	Name of the configMap used to support HDFS configurations when using HDFS as Jindo's UFS. The configMap must be in the same namespace with the JindoRuntime. The configMap should contain user-specific HDFS conf files in it. For now, only "hdfs-site.xml" and "core-site.xml" are supported. It must take the filename of the conf file as the key and content of the file as the value.		Optional: {}
secret string
labels object (keys:string, values:string)	Labels will be added on all the JindoFS pods. DEPRECATED: this is a deprecated field. Please use PodMetadata.Labels instead. Note: this field is set to be exclusive with PodMetadata.Labels		Optional: {}
podMetadata PodMetadata	PodMetadata defines labels and annotations that will be propagated to all Jindo's fuse pods		Optional: {}
logConfig object (keys:string, values:string)			Optional: {}
networkmode NetworkMode	Whether to use hostnetwork or not		Enum: [HostNetwork ContainerNetwork] Optional: {}
cleanCachePolicy CleanCachePolicy	CleanCachePolicy defines cleanCache Policy		Optional: {}
volumes Volume array	Volumes is the list of Kubernetes volumes that can be mounted by the jindo runtime components and/or fuses.		Optional: {}
imagePullSecrets LocalObjectReference array	ImagePullSecrets that will be used to pull images		Optional: {}

JobProcessor

Appears in:

• Processor

Field	Description	Default	Validation
podSpec PodSpec	PodSpec defines Pod specification of the DataProcess job.		Optional: {}

JuiceFSCompTemplateSpec

JuiceFSCompTemplateSpec is a description of the JuiceFS components

Appears in:

• JuiceFSRuntimeSpec

Field	Description	Default	Validation
replicas integer	Replicas is the desired number of replicas of the given template. If unspecified, defaults to 1. replicas is the min replicas of dataset in the cluster		Minimum: 1 Optional: {}
ports ContainerPort array	Ports used by JuiceFS		Optional: {}
resources ResourceRequirements	Resources that will be requested by the JuiceFS component.		Optional: {}
options object (keys:string, values:string)	Options
env EnvVar array	Environment variables that will be used by JuiceFS component.
enabled boolean	Enabled or Disabled for the components.		Optional: {}
nodeSelector object (keys:string, values:string)	NodeSelector is a selector		Optional: {}
volumeMounts VolumeMount array	VolumeMounts specifies the volumes listed in ".spec.volumes" to mount into runtime component's filesystem.		Optional: {}
podMetadata PodMetadata	PodMetadata defines labels and annotations that will be propagated to JuiceFs's pods.		Optional: {}
networkMode NetworkMode	Whether to use hostnetwork or not		Enum: [HostNetwork ContainerNetwork] Optional: {}

JuiceFSFuseSpec

Appears in:

• JuiceFSRuntimeSpec

Field	Description	Default	Validation
image string	Image for JuiceFS fuse
imageTag string	Image for JuiceFS fuse
imagePullPolicy string	One of the three policies: Always, IfNotPresent, Never
env EnvVar array	Environment variables that will be used by JuiceFS Fuse
resources ResourceRequirements	Resources that will be requested by JuiceFS Fuse.
options object (keys:string, values:string)	Options mount options that fuse pod will use		Optional: {}
nodeSelector object (keys:string, values:string)	NodeSelector is a selector which must be true for the fuse client to fit on a node, this option only effect when global is enabled		Optional: {}
volumeMounts VolumeMount array	VolumeMounts specifies the volumes listed in ".spec.volumes" to mount into runtime component's filesystem.		Optional: {}
cleanPolicy FuseCleanPolicy	CleanPolicy decides when to clean Juicefs Fuse pods. Currently Fluid supports three policies: OnDemand, OnRuntimeDeleted and OnFuseChangedCleanPolicy OnDemand cleans fuse pod once the fuse pod on some node is not needed OnRuntimeDeleted cleans fuse pod only when the cache runtime is deleted OnFuseChangedCleanPolicy cleans fuse pod once the fuse pod on some node is not needed and the fuse in runtime is updated Defaults to OnRuntimeDeleted		Optional: {}
podMetadata PodMetadata	PodMetadata defines labels and annotations that will be propagated to JuiceFs's pods.		Optional: {}
networkMode NetworkMode	Whether to use hostnetwork or not		Enum: [HostNetwork ContainerNetwork] Optional: {}

JuiceFSRuntime

JuiceFSRuntime is the Schema for the juicefsruntimes API

Field	Description	Default	Validation
apiVersion string	data.fluid.io/v1alpha1
kind string	JuiceFSRuntime
metadata ObjectMeta	Refer to Kubernetes API documentation for fields of metadata.
spec JuiceFSRuntimeSpec

JuiceFSRuntimeSpec

JuiceFSRuntimeSpec defines the desired state of JuiceFSRuntime

Appears in:

• JuiceFSRuntime

Field	Description	Default	Validation
juicefsVersion VersionSpec	The version information that instructs fluid to orchestrate a particular version of JuiceFS.
initUsers InitUsersSpec	The spec of init users
master JuiceFSCompTemplateSpec	The component spec of JuiceFS master
worker JuiceFSCompTemplateSpec	The component spec of JuiceFS worker
jobWorker JuiceFSCompTemplateSpec	The component spec of JuiceFS job Worker
fuse JuiceFSFuseSpec	Desired state for JuiceFS Fuse
tieredstore TieredStore	Tiered storage used by JuiceFS
configs string	Configs of JuiceFS
replicas integer	The replicas of the worker, need to be specified
runAs User	Manage the user to run Juicefs Runtime
disablePrometheus boolean	Disable monitoring for JuiceFS Runtime Prometheus is enabled by default		Optional: {}
volumes Volume array	Volumes is the list of Kubernetes volumes that can be mounted by the alluxio runtime components and/or fuses.		Optional: {}
podMetadata PodMetadata	PodMetadata defines labels and annotations that will be propagated to JuiceFs's pods.		Optional: {}
management RuntimeManagement	RuntimeManagement defines policies when managing the runtime		Optional: {}

Level

Level describes configurations a tier needs.
Refer to Configuring Tiered Storage for more info

Appears in:

• TieredStore

Field	Description	Default	Validation
mediumtype MediumType	Medium Type of the tier. One of the three types: MEM, SSD, HDD		Enum: [MEM SSD HDD] Required: {}
volumeType VolumeType	VolumeType is the volume type of the tier. Should be one of the three types: hostPath, emptyDir and volumeTemplate. If not set, defaults to hostPath.	hostPath	Enum: [hostPath emptyDir] Optional: {}
path string	File paths to be used for the tier. Multiple paths are supported. Multiple paths should be separated with comma. For example: "/mnt/cache1,/mnt/cache2".		MinLength: 1 Optional: {}
quota Quantity	Quota for the whole tier. (e.g. 100Gi) Please note that if there're multiple paths used for this tierstore, the quota will be equally divided into these paths. If you'd like to set quota for each, path, see QuotaList for more information.		Optional: {}
quotaList string	QuotaList are quotas used to set quota on multiple paths. Quotas should be separated with comma. Quotas in this list will be set to paths with the same order in Path. For example, with Path defined with "/mnt/cache1,/mnt/cache2" and QuotaList set to "100Gi, 50Gi", then we get 100GiB cache storage under "/mnt/cache1" and 50GiB under "/mnt/cache2". Also note that num of quotas must be consistent with the num of paths defined in Path.		Pattern: ^((\+|-)?(([0-9]+(\.[0-9]*)?)|(\.[0-9]+))(([KMGTPE]i)|[numkMGTPE]|([eE](\+|-)?(([0-9]+(\.[0-9]*)?)|(\.[0-9]+)))),)+((\+|-)?(([0-9]+(\.[0-9]*)?)|(\.[0-9]+))(([KMGTPE]i)|[numkMGTPE]|([eE](\+|-)?(([0-9]+(\.[0-9]*)?)|(\.[0-9]+))))?)$ Optional: {}
high string	Ratio of high watermark of the tier (e.g. 0.9)
low string	Ratio of low watermark of the tier (e.g. 0.7)

MasterSpec

MasterSpec defines the configurations for Vineyard Master component which is also regarded as the Etcd component in Vineyard. For more info about Vineyard, refer to Vineyard

Appears in:

• VineyardRuntimeSpec

Field	Description	Default	Validation
replicas integer	The replicas of Vineyard component. If not specified, defaults to 1. For worker, the replicas should not be greater than the number of nodes in the cluster		Minimum: 1 Optional: {}
image string	The image of Vineyard component. For Master, the default image is registry.aliyuncs.com/vineyard/vineyardd For Worker, the default image is registry.aliyuncs.com/vineyard/vineyardd The default container registry is docker.io, you can change it by setting the image field		Optional: {}
imageTag string	The image tag of Vineyard component. For Master, the default image tag is v0.22.2. For Worker, the default image tag is v0.22.2.		Optional: {}
imagePullPolicy string	The image pull policy of Vineyard component. Default is IfNotPresent.		Optional: {}
nodeSelector object (keys:string, values:string)	NodeSelector is a selector to choose which nodes to launch the Vineyard component. E,g. {"disktype": "ssd"}		Optional: {}
ports object (keys:string, values:integer)	Ports used by Vineyard component. For Master, the default client port is 2379 and peer port is 2380. For Worker, the default rpc port is 9600 and the default exporter port is 9144.		Optional: {}
env object (keys:string, values:string)	Environment variables that will be used by Vineyard component. For Master, refer to Etcd Configuration for more info Default is not set.		Optional: {}
options object (keys:string, values:string)	Configurable options for Vineyard component. For Master, there is no configurable options. For Worker, support the following options. vineyardd.reserve.memory: (Bool) where to reserve memory for vineyardd If set to true, the memory quota will be counted to the vineyardd rather than the application. etcd.prefix: (String) the prefix of etcd key for vineyard objects wait.etcd.timeout: (String) the timeout period before waiting the etcd to be ready, in seconds Default value is as follows. vineyardd.reserve.memory: "true" etcd.prefix: "/vineyard" wait.etcd.timeout: "120"		Optional: {}
resources ResourceRequirements	Resources contains the resource requirements and limits for the Vineyard component. Default is not set. For Worker, when the options contains vineyardd.reserve.memory=true, the resources.request.memory for worker should be greater than tieredstore.levels[0].quota(aka vineyardd shared memory)		Optional: {}
volumeMounts VolumeMount array	VolumeMounts specifies the volumes listed in ".spec.volumes" to mount into the vineyard runtime component's filesystem. It is useful for specifying a persistent storage. Default is not set.		Optional: {}
podMetadata PodMetadata	PodMetadata defines labels and annotations that will be propagated to Vineyard's pods including Master and Worker. Default is not set.		Optional: {}
networkMode NetworkMode	Whether to use hostnetwork or not Default is HostNetwork		Enum: [HostNetwork ContainerNetwork] Optional: {}
endpoint ExternalEndpointSpec	ExternalEndpoint defines the configurations for external etcd cluster Default is not set If set, the Vineyard Master component will not be deployed, which means the Vineyard Worker component will use an external Etcd cluster. E,g. endpoint: uri: "etcd-svc.etcd-namespace.svc.cluster.local:2379" encryptOptions: - name: access-key valueFrom: secretKeyRef: name: etcd-secret key: accesskey		Optional: {}

MediumSource

MediumSource describes the storage medium type for tiered store. Only one of its members may be specified.

Appears in:

• RuntimeTieredStoreLevel

Field	Description	Default	Validation
processMemory ProcessMemoryMediumSource	ProcessMemory indicates that process memory should be used as the storage medium. The cache will be stored in the process's memory space.		Optional: {}
volume VolumeMediumSource	Volume indicates that a Kubernetes volume should be used as the storage medium. Supported volume types include hostPath, emptyDir, and ephemeral volumes.		Optional: {}

MetadataSyncPolicy

MetadataSyncPolicy defines policies when syncing metadata

Appears in:

• RuntimeManagement

Field	Description	Default	Validation
autoSync boolean	AutoSync enables automatic metadata sync when setting up a runtime. If not set, it defaults to true.		Optional: {}

Mount

Mount describes a mounting.
Refer to Alluxio Storage Integrations for more info

Appears in:

• DatasetSpec
• DatasetStatus
• MountPointStatus
• RuntimeStatus

Field	Description	Default	Validation
mountPoint string	MountPoint is the mount point of source.		MinLength: 5 Required: {}
options object (keys:string, values:string)	The Mount Options. Refer to Mount Options. The option has Prefix 'fs.' And you can Learn more from The Storage Integrations		Optional: {}
name string	The name of mount		MinLength: 0 Optional: {}
path string	The path of mount, if not set will be /{Name}		Optional: {}
readOnly boolean	Optional: Defaults to false (read-write).		Optional: {}
shared boolean	Optional: Defaults to false (shared).		Optional: {}
encryptOptions EncryptOption array	The secret information		Optional: {}

MountPointStatus

MountPointStatus describes the status of a single mount point in the dataset

Appears in:

• CacheRuntimeStatus

Field	Description	Default	Validation
mount Mount	Mount contains the mount point configuration from the bound dataset. This includes the remote path, mount options, and other mount-specific settings.
mountTime Time	MountTime is the timestamp of the last successful mount operation. If MountTime is earlier than the master component's start time, a remount will be required.		Optional: {}

NetworkMode

Underlying type: string

Appears in:

• AlluxioCompTemplateSpec
• AlluxioFuseSpec
• EFCCompTemplateSpec
• EFCFuseSpec
• JindoRuntimeSpec
• JuiceFSCompTemplateSpec
• JuiceFSFuseSpec
• MasterSpec
• ThinCompTemplateSpec
• ThinFuseSpec
• VineyardClientSocketSpec
• VineyardCompTemplateSpec

Field	Description
HostNetwork
ContainerNetwork
``	DefaultNetworkMode is Host

NodePublishSecretPolicy

Underlying type: string

Appears in:

• ThinRuntimeProfileSpec

Field	Description
NotMountNodePublishSecret
MountNodePublishSecretIfExists
CopyNodePublishSecretAndMountIfNotExists

OSAdvise

OSAdvise is a description of choices to have optimization on specific operating system

Appears in:

• EFCRuntimeSpec

Field	Description	Default	Validation
osVersion string	Specific operating system version that can have optimization.		Optional: {}
enabled boolean	Enable operating system optimization not enabled by default.		Optional: {}

ObjectRef

Appears in:

• AffinityStrategy
• OperationRef

Field	Description	Default	Validation
apiVersion string	API version of the referent operation		Optional: {}
kind string	Kind specifies the type of the referent operation		Enum: [DataLoad DataBackup DataMigrate DataProcess] Required: {}
name string	Name specifies the name of the referent operation		Required: {}
namespace string	Namespace specifies the namespace of the referent operation.		Optional: {}

OperationRef

Appears in:

• DataBackupSpec
• DataLoadSpec
• DataMigrateSpec
• DataProcessSpec

Field	Description	Default	Validation
apiVersion string	API version of the referent operation		Optional: {}
kind string	Kind specifies the type of the referent operation		Enum: [DataLoad DataBackup DataMigrate DataProcess] Required: {}
name string	Name specifies the name of the referent operation		Required: {}
namespace string	Namespace specifies the namespace of the referent operation.		Optional: {}
affinityStrategy AffinityStrategy	AffinityStrategy specifies the pod affinity strategy with the referent operation.		Optional: {}

PlacementMode

Underlying type: string

Appears in:

• DatasetSpec

Field	Description
Exclusive
Shared
``	DefaultMode is exclusive

PodMetadata

PodMetadata defines subgroup properties of metav1.ObjectMeta

Appears in:

• AlluxioCompTemplateSpec
• AlluxioFuseSpec
• AlluxioRuntimeSpec
• CacheRuntimeClientSpec
• CacheRuntimeMasterSpec
• CacheRuntimeSpec
• CacheRuntimeWorkerSpec
• DataLoadSpec
• DataMigrateSpec
• EFCCompTemplateSpec
• EFCFuseSpec
• EFCRuntimeSpec
• JindoCompTemplateSpec
• JindoFuseSpec
• JindoRuntimeSpec
• JuiceFSCompTemplateSpec
• JuiceFSFuseSpec
• JuiceFSRuntimeSpec
• MasterSpec
• Metadata
• Processor
• RuntimeComponentCommonSpec
• ThinFuseSpec
• VineyardClientSocketSpec
• VineyardCompTemplateSpec
• VineyardRuntimeSpec

Field	Description	Default	Validation
labels object (keys:string, values:string)	Labels are labels of pod specification
annotations object (keys:string, values:string)	Annotations are annotations of pod specification

Policy

Underlying type: string

Appears in:

• DataLoadSpec
• DataMigrateSpec

Field	Description
Once	Once run data migrate once, default policy is Once
Cron	Cron run data migrate by cron
OnEvent	OnEvent run data migrate when event occurs

Prefer

Prefer defines the label key and weight for generating a PreferredSchedulingTerm.

Appears in:

• AffinityStrategy

Field	Description	Default	Validation
name string
weight integer

ProcessMemoryMediumSource

ProcessMemoryMediumSource describes process memory as a storage medium. When specified, cache data will be stored in the process's memory space, and the quota will be added to the container's resource requests and limits.

Appears in:

• MediumSource

Processor

Processor defines the actual processor for DataProcess. Processor can be either of a Job or a Shell script.

Appears in:

• DataProcessSpec

Field	Description	Default	Validation
serviceAccountName string	ServiceAccountName defiens the serviceAccountName of the container		Optional: {}
podMetadata PodMetadata	PodMetadata defines labels and annotations on the processor pod.		Optional: {}
job JobProcessor	Job represents a processor which runs DataProcess as a job.		Optional: {}
script ScriptProcessor	Shell represents a processor which executes shell script

Require

Require defines the label key for generating a NodeSelectorTerm.

Appears in:

• AffinityStrategy

Field	Description	Default	Validation
name string

Runtime

Runtime describes a runtime to be used to support dataset

Appears in:

• DatasetSpec
• DatasetStatus

Field	Description	Default	Validation
name string	Name of the runtime object
namespace string	Namespace of the runtime object
category Category	Category the runtime object belongs to (e.g. Accelerate)
type string	Runtime object's type (e.g. Alluxio)
masterReplicas integer	Runtime master replicas

RuntimeComponentCommonSpec

RuntimeComponentCommonSpec describes the common configuration for CacheRuntime components

Appears in:

• CacheRuntimeClientSpec
• CacheRuntimeMasterSpec
• CacheRuntimeWorkerSpec

Field	Description	Default	Validation
disabled boolean	Disabled indicates whether the component should be disabled. If set to true, the component will not be created.		Optional: {}
runtimeVersion VersionSpec	RuntimeVersion is the version information that instructs Fluid to orchestrate a particular version of the runtime.		Optional: {}
options object (keys:string, values:string)	Options is a set of key-value pairs that provide additional configuration for the cache system.		Optional: {}
podMetadata PodMetadata	PodMetadata contains labels and annotations that will be propagated to the component's pods.		Optional: {}
resources ResourceRequirements	Resources describes the compute resource requirements. More info: https://kubernetes.io/docs/concepts/configuration/manage-resources-containers/		Optional: {}
env EnvVar array	Env is a list of environment variables to set in the container.		Optional: {}
volumeMounts VolumeMount array	VolumeMounts are the volumes to mount into the container's filesystem. Cannot be updated.		Optional: {}
args string array	Args are arguments to the entrypoint. The container image's CMD is used if this is not provided.		Optional: {}
nodeSelector object (keys:string, values:string)	NodeSelector is a selector which must be true for the pod to fit on a node. Selector which must match a node's labels for the pod to be scheduled on that node. More info: https://kubernetes.io/docs/concepts/configuration/assign-pod-node/		Optional: {}
tolerations Toleration array	Tolerations are the pod's tolerations. If specified, the pod's tolerations. More info: https://kubernetes.io/docs/concepts/scheduling-eviction/taint-and-toleration/		Optional: {}

RuntimeComponentDefinition

RuntimeComponentDefinition defines the configuration for a CacheRuntime component

Appears in:

• RuntimeTopology

Field	Description	Default	Validation
workloadType TypeMeta	WorkloadType is the default workload type of the component		Optional: {}
options object (keys:string, values:string)	Options is a set of key-value pairs that provide additional configuration for the component		Optional: {}
template PodTemplateSpec	Template describes the pods that will be created. The template follows the standard PodTemplateSpec from Kubernetes core.		Optional: {}
service RuntimeComponentService	Service is the service configuration for the component		Optional: {}
dependencies RuntimeComponentDependencies	Dependencies specifies the dependencies required by the component		Optional: {}

RuntimeComponentDependencies

RuntimeComponentDependencies defines the dependencies required by a CacheRuntime component

Appears in:

• RuntimeComponentDefinition

Field	Description	Default	Validation
encryptOption EncryptOptionComponentDependency	EncryptOption is the configuration for encrypt option secret mount		Optional: {}
extraResources ExtraResourcesComponentDependency	ExtraResources specifies the usage of extra resources such as ConfigMaps		Optional: {}

RuntimeComponentService

RuntimeComponentService describes the service configuration for a CacheRuntime component

Appears in:

• RuntimeComponentDefinition

Field	Description	Default	Validation
headless HeadlessRuntimeComponentService	Headless enables a headless service for the component. A headless service does not allocate a cluster IP and allows direct pod-to-pod communication.		Optional: {}

RuntimeComponentStatus

RuntimeComponentStatus describes the observed state of a runtime component. It follows the standard Kubernetes pattern for tracking workload status.

Appears in:

• CacheRuntimeStatus
• RuntimeComponentStatusCollection

Field	Description	Default	Validation
phase RuntimePhase	Phase is the current lifecycle phase of the component. Known phases include: "Pending", "Running", "Failed", etc.		Required: {}
reason string	Reason is a brief, machine-readable string that gives the reason for the current phase. This is useful for understanding why a component is in a particular state.		Optional: {}
message string	Message is a human-readable message indicating details about the component's status.		Optional: {}
readyReplicas integer	ReadyReplicas is the number of pods with a Ready condition.		Optional: {}
currentReplicas integer	CurrentReplicas is the current number of replicas running for the component		Optional: {}
desiredReplicas integer	DesiredReplicas is the desired number of replicas as specified in the spec.		Optional: {}
unavailableReplicas integer	UnavailableReplicas is the number of pods that are not available. A pod is considered unavailable if it is not ready or has been terminated.		Optional: {}
availableReplicas integer	AvailableReplicas is the number of pods that are available and ready to serve requests. This count includes only pods that have passed readiness checks.		Optional: {}

RuntimeComponentStatusCollection

RuntimeComponentStatusCollection describes the status of all runtime components. It provides a unified view of master, worker, and client component states.

Appears in:

• CacheRuntimeStatus

Field	Description	Default	Validation
master RuntimeComponentStatus	Master is the observed state of the master component.		Optional: {}
worker RuntimeComponentStatus	Worker is the observed state of the worker component.		Optional: {}
client RuntimeComponentStatus	Client is the observed state of the client (FUSE) component.		Optional: {}

RuntimeCondition

Condition describes the state of the cache at a certain point.

Appears in:

• CacheRuntimeStatus
• RuntimeStatus

Field	Description	Default	Validation
type RuntimeConditionType	Type of cache condition.
reason string	The reason for the condition's last transition.
message string	A human readable message indicating details about the transition.
lastProbeTime Time	The last time this condition was updated.
lastTransitionTime Time	Last time the condition transitioned from one status to another.

RuntimeConditionType

Underlying type: string

RuntimeConditionType indicates valid conditions type of a runtime

Appears in:

• RuntimeCondition

Field	Description
MasterInitialized	RuntimeMasterInitialized means the master of runtime is initialized
MasterReady	RuntimeMasterReady means the master of runtime is ready
WorkersInitialized	RuntimeWorkersInitialized means the workers of runtime are initialized
WorkersReady	RuntimeWorkersReady means the workers of runtime are ready
WorkersScaledIn	RuntimeWorkerScaledIn means the workers of runtime just scaled in
WorkersScaledOut	RuntimeWorkerScaledIn means the workers of runtime just scaled out
FusesInitialized	RuntimeFusesInitialized means the fuses of runtime are initialized
FusesReady	RuntimeFusesReady means the fuses of runtime are ready
FusesScaledIn	RuntimeFusesScaledIn means the fuses of runtime just scaled in
FusesScaledOut	RuntimeFusesScaledOut means the fuses of runtime just scaled out

RuntimeExtraResources

RuntimeExtraResources defines the extra resources for CacheRuntime

Appears in:

• CacheRuntimeClass

Field	Description	Default	Validation
configMaps ConfigMapRuntimeExtraResource array	ConfigMaps is a list of ConfigMaps that will be created in the runtime's namespace. These ConfigMaps can be referenced and mounted by runtime components.		Optional: {}

RuntimeManagement

RuntimeManagement defines suggestions for runtime controllers to manage the runtime

Appears in:

• AlluxioRuntimeSpec
• JuiceFSRuntimeSpec
• ThinRuntimeSpec

Field	Description	Default	Validation
cleanCachePolicy CleanCachePolicy	CleanCachePolicy defines the policy of cleaning cache when shutting down the runtime		Optional: {}
metadataSyncPolicy MetadataSyncPolicy	MetadataSyncPolicy defines the policy of syncing metadata when setting up the runtime. If not set,		Optional: {}

RuntimePhase

Underlying type: string

Appears in:

• RuntimeComponentStatus
• RuntimeStatus

Field	Description
``
NotReady
PartialReady
Ready

RuntimeTieredStore

RuntimeTieredStore describes the tiered storage configuration for cache runtime

Appears in:

• CacheRuntimeClientSpec
• CacheRuntimeWorkerSpec

Field	Description	Default	Validation
levels RuntimeTieredStoreLevel array	Levels is the list of cache storage tiers from highest priority to lowest. Each tier can use different storage media (e.g., memory, SSD, HDD).		Optional: {}

RuntimeTieredStoreLevel

RuntimeTieredStoreLevel describes the configuration for a single tier in the tiered storage

Appears in:

• RuntimeTieredStore

Field	Description	Default	Validation
medium MediumSource	Medium describes the storage medium type for this tier. Supported types include process memory and various volume types.		Optional: {}
path string array	Path is a list of file paths to be used for the cache tier. Multiple paths can be specified to distribute cache across different mount points. For example: ["/mnt/cache1", "/mnt/cache2"].		MinItems: 1 Optional: {}
quota Quantity array	Quota is a list of storage quotas for each path in the tier. The length of Quota should match the length of Path. Each quota corresponds to the path at the same index. For example: ["100Gi", "50Gi"] allocates 100GiB to the first path and 50GiB to the second path.		Optional: {}
high string	High is the ratio of high watermark of the tier (e.g., "0.9"). When cache usage exceeds this ratio, eviction will be triggered.		Optional: {}
low string	Low is the ratio of low watermark of the tier (e.g., "0.7"). Eviction will continue until cache usage falls below this ratio.		Optional: {}

RuntimeTopology

RuntimeTopology defines the topology structure of CacheRuntime components

Appears in:

• CacheRuntimeClass

Field	Description	Default	Validation
master RuntimeComponentDefinition	Master is the configuration for master component		Optional: {}
worker RuntimeComponentDefinition	Worker is the configuration for worker component		Optional: {}
client RuntimeComponentDefinition	Client is the configuration for client component		Optional: {}

ScriptProcessor

Appears in:

• Processor

Field	Description	Default	Validation
image string	Image (e.g. alluxio/alluxio)
imageTag string	Image tag (e.g. 2.3.0-SNAPSHOT)
imagePullPolicy string	One of the three policies: Always, IfNotPresent, Never
restartPolicy RestartPolicy	RestartPolicy specifies the processor job's restart policy. Only "Never", "OnFailure" is allowed.	Never	Enum: [Never OnFailure] Optional: {}
command string array	Entrypoint command for ScriptProcessor.		Optional: {}
source string	Script source for ScriptProcessor		Required: {}
env EnvVar array	List of environment variables to set in the container.		Optional: {}
volumeMounts VolumeMount array	Pod volumes to mount into the container's filesystem.		Optional: {}
volumes Volume array	List of volumes that can be mounted by containers belonging to the pod.		Optional: {}
resources ResourceRequirements	Resources that will be requested by the DataProcess job.		Optional: {}

SecretKeySelector

Appears in:

• EncryptOptionSource

Field	Description	Default	Validation
name string	The name of required secret		Required: {}
key string	The required key in the secret		Optional: {}

TargetDataset

TargetDataset defines the target dataset of the DataLoad

Appears in:

• DataLoadSpec
• TargetDatasetWithMountPath

Field	Description	Default	Validation
name string	Name defines name of the target dataset
namespace string	Namespace defines namespace of the target dataset

TargetDatasetWithMountPath

TargetDataset defines which dataset will be processed by DataProcess. Under the hood, the dataset's pvc will be mounted to the given mountPath of the DataProcess's containers.

Appears in:

• DataProcessSpec

Field	Description	Default	Validation
name string	Name defines name of the target dataset
namespace string	Namespace defines namespace of the target dataset
mountPath string	MountPath defines where the Dataset should be mounted in DataProcess's containers.		Required: {}
subPath string	SubPath defines subpath of the target dataset to mount.		Optional: {}

TargetPath

TargetPath defines the target path of the DataLoad

Appears in:

• DataLoadSpec

Field	Description	Default	Validation
path string	Path defines path to be load
replicas integer	Replicas defines how many replicas will be loaded

ThinCompTemplateSpec

ThinCompTemplateSpec is a description of the thinRuntime components

Appears in:

• ThinRuntimeProfileSpec
• ThinRuntimeSpec

Field	Description	Default	Validation
image string	Image for thinRuntime fuse
imageTag string	Image for thinRuntime fuse
imagePullPolicy string	One of the three policies: Always, IfNotPresent, Never
imagePullSecrets LocalObjectReference array	ImagePullSecrets that will be used to pull images		Optional: {}
replicas integer	Replicas is the desired number of replicas of the given template. If unspecified, defaults to 1. replicas is the min replicas of dataset in the cluster		Minimum: 1 Optional: {}
ports ContainerPort array	Ports used thinRuntime		Optional: {}
resources ResourceRequirements	Resources that will be requested by thinRuntime component.		Optional: {}
env EnvVar array	Environment variables that will be used by thinRuntime component.
enabled boolean	Enabled or Disabled for the components.		Optional: {}
nodeSelector object (keys:string, values:string)	NodeSelector is a selector		Optional: {}
volumeMounts VolumeMount array	VolumeMounts specifies the volumes listed in ".spec.volumes" to mount into runtime component's filesystem.		Optional: {}
livenessProbe Probe	livenessProbe of thin fuse pod		Optional: {}
readinessProbe Probe	readinessProbe of thin fuse pod		Optional: {}
networkMode NetworkMode	Whether to use hostnetwork or not		Enum: [HostNetwork ContainerNetwork] Optional: {}

ThinFuseSpec

Appears in:

• ThinRuntimeProfileSpec
• ThinRuntimeSpec

Field	Description	Default	Validation
image string	Image for thinRuntime fuse
imageTag string	Image for thinRuntime fuse
imagePullPolicy string	One of the three policies: Always, IfNotPresent, Never
imagePullSecrets LocalObjectReference array	ImagePullSecrets that will be used to pull images		Optional: {}
ports ContainerPort array	Ports used thinRuntime		Optional: {}
env EnvVar array	Environment variables that will be used by thinRuntime Fuse
command string array	Command that will be passed to thinRuntime Fuse
args string array	Arguments that will be passed to thinRuntime Fuse
options object (keys:string, values:string)	Options configurable options of FUSE client, performance parameters usually. will be merged with Dataset.spec.mounts.options into fuse pod.
resources ResourceRequirements	Resources that will be requested by thinRuntime Fuse.
nodeSelector object (keys:string, values:string)	NodeSelector is a selector which must be true for the fuse client to fit on a node, this option only effect when global is enabled		Optional: {}
cleanPolicy FuseCleanPolicy	CleanPolicy decides when to clean thinRuntime Fuse pods. Currently Fluid supports two policies: OnDemand and OnRuntimeDeleted OnDemand cleans fuse pod once the fuse pod on some node is not needed OnRuntimeDeleted cleans fuse pod only when the cache runtime is deleted Defaults to OnDemand		Optional: {}
networkMode NetworkMode	Whether to use hostnetwork or not		Enum: [HostNetwork ContainerNetwork] Optional: {}
livenessProbe Probe	livenessProbe of thin fuse pod		Optional: {}
readinessProbe Probe	readinessProbe of thin fuse pod		Optional: {}
volumeMounts VolumeMount array	VolumeMounts specifies the volumes listed in ".spec.volumes" to mount into the thinruntime component's filesystem.		Optional: {}
lifecycle Lifecycle	Lifecycle describes actions that the management system should take in response to container lifecycle events.
podMetadata PodMetadata	PodMetadata defines labels and annotations that will be propagated to ThinRuntime's FUSE pods.		Optional: {}

ThinRuntime

ThinRuntime is the Schema for the thinruntimes API

Field	Description	Default	Validation
apiVersion string	data.fluid.io/v1alpha1
kind string	ThinRuntime
metadata ObjectMeta	Refer to Kubernetes API documentation for fields of metadata.
spec ThinRuntimeSpec

ThinRuntimeProfile

ThinRuntimeProfile is the Schema for the ThinRuntimeProfiles API

Field	Description	Default	Validation
apiVersion string	data.fluid.io/v1alpha1
kind string	ThinRuntimeProfile
metadata ObjectMeta	Refer to Kubernetes API documentation for fields of metadata.
spec ThinRuntimeProfileSpec

ThinRuntimeProfileSpec

ThinRuntimeProfileSpec defines the desired state of ThinRuntimeProfile

Appears in:

• ThinRuntimeProfile

Field	Description	Default	Validation
fileSystemType string	file system of thinRuntime		Required: {}
imagePullSecrets LocalObjectReference array	ImagePullSecrets that will be used to pull images		Optional: {}
worker ThinCompTemplateSpec	The component spec of worker
fuse ThinFuseSpec	The component spec of thinRuntime
volumes Volume array	Volumes is the list of Kubernetes volumes that can be mounted by runtime components and/or fuses.		Optional: {}
nodePublishSecretPolicy NodePublishSecretPolicy	NodePublishSecretPolicy describes the policy to decide which to do with node publish secret when mounting an existing persistent volume.	MountNodePublishSecretIfExists	Enum: [NotMountNodePublishSecret MountNodePublishSecretIfExists CopyNodePublishSecretAndMountIfNotExists]

ThinRuntimeSpec

ThinRuntimeSpec defines the desired state of ThinRuntime

Appears in:

• ThinRuntime

Field	Description	Default	Validation
profileName string	The specific runtime profile name, empty value is used for handling datasets which mount another dataset
worker ThinCompTemplateSpec	The component spec of worker
fuse ThinFuseSpec	The component spec of thinRuntime
tieredstore TieredStore	Tiered storage
replicas integer	The replicas of the worker, need to be specified
runAs User	Manage the user to run Runtime
disablePrometheus boolean	Disable monitoring for Runtime Prometheus is enabled by default		Optional: {}
volumes Volume array	Volumes is the list of Kubernetes volumes that can be mounted by runtime components and/or fuses.		Optional: {}
management RuntimeManagement	RuntimeManagement defines policies when managing the runtime		Optional: {}
imagePullSecrets LocalObjectReference array	ImagePullSecrets that will be used to pull images		Optional: {}

TieredStore

TieredStore is a description of the tiered store

Appears in:

• AlluxioRuntimeSpec
• EFCRuntimeSpec
• GooseFSRuntimeSpec
• JindoRuntimeSpec
• JuiceFSRuntimeSpec
• ThinRuntimeSpec
• VineyardRuntimeSpec

Field	Description	Default	Validation
levels Level array	configurations for multiple tiers

User

User explains the user and group to run a Container

Appears in:

• AlluxioRuntimeSpec
• DataBackupSpec
• DatasetSpec
• GooseFSRuntimeSpec
• JindoRuntimeSpec
• JuiceFSRuntimeSpec
• ThinRuntimeSpec

Field	Description	Default	Validation
uid integer	The uid to run the alluxio runtime
gid integer	The gid to run the alluxio runtime
user string	The user name to run the alluxio runtime
group string	The group name to run the alluxio runtime

VersionSpec

VersionSpec represents the settings for the version that fluid is orchestrating.

Appears in:

• AlluxioRuntimeSpec
• CacheRuntimeClientSpec
• CacheRuntimeMasterSpec
• CacheRuntimeWorkerSpec
• DataMigrateSpec
• EFCCompTemplateSpec
• EFCFuseSpec
• GooseFSRuntimeSpec
• InitFuseSpec
• JindoRuntimeSpec
• JuiceFSRuntimeSpec
• RuntimeComponentCommonSpec
• ScriptProcessor

Field	Description	Default	Validation
image string	Image (e.g. alluxio/alluxio)
imageTag string	Image tag (e.g. 2.3.0-SNAPSHOT)
imagePullPolicy string	One of the three policies: Always, IfNotPresent, Never

VineyardClientSocketSpec

VineyardClientSocketSpec holds the configurations for vineyard client socket

Appears in:

• VineyardRuntimeSpec

Field	Description	Default	Validation
image string	Image for Vineyard Fuse Default is registry.aliyuncs.com/vineyard/vineyard-fluid-fuse		Optional: {}
imageTag string	Image Tag for Vineyard Fuse Default is v0.22.2		Optional: {}
imagePullPolicy string	Image pull policy for Vineyard Fuse Default is IfNotPresent Available values are Always, IfNotPresent, Never		Optional: {}
env object (keys:string, values:string)	Environment variables that will be used by Vineyard Fuse. Default is not set.		Optional: {}
cleanPolicy FuseCleanPolicy	CleanPolicy decides when to clean Vineyard Fuse pods. Currently Fluid supports two policies: OnDemand and OnRuntimeDeleted OnDemand cleans fuse pod once th fuse pod on some node is not needed OnRuntimeDeleted cleans fuse pod only when the cache runtime is deleted Defaults to OnRuntimeDeleted		Optional: {}
resources ResourceRequirements	Resources contains the resource requirements and limits for the Vineyard Fuse. Default is not set.		Optional: {}
networkMode NetworkMode	Whether to use hostnetwork or not Default is HostNetwork		Enum: [HostNetwork ContainerNetwork] Optional: {}
podMetadata PodMetadata	PodMetadata defines labels and annotations that will be propagated to Vineyard's pods.		Optional: {}
options object (keys:string, values:string)	Options for configuring vineyardd parameters. Supported options are as follows. reserve_memory: (Bool) Whether to reserving enough physical memory pages for vineyardd. Default is true. allocator: (String) The allocator used by vineyardd, could be "dlmalloc" or "mimalloc". Default is "dlmalloc". compression: (Bool) Compress before migration or spilling. Default is true. coredump: (Bool) Enable coredump core dump when been aborted. Default is false. meta_timeout: (Int) Timeout period before waiting the metadata service to be ready, in seconds Default is 60. etcd_endpoint: (String) The endpoint of etcd. Default is same as the etcd endpoint of vineyard worker. etcd_prefix: (String) Metadata path prefix in etcd. Default is "/vineyard". size: (String) shared memory size for vineyardd. 1024M, 1024000, 1G, or 1Gi. Default is "0", which means no cache. When the size is not set to "0", it should be greater than the 2048 bytes(2K). spill_path: (String) Path to spill temporary files, if not set, spilling will be disabled. Default is "". spill_lower_rate: (Double) The lower rate of memory usage to trigger spilling. Default is 0.3. spill_upper_rate: (Double) The upper rate of memory usage to stop spilling. Default is 0.8. Default is as follows. fuse: options: size: "0" etcd_endpoint: "http://{{Name}}-master-0.{{Name}}-master.{{Namespace}}:{{EtcdClientPort}}" etcd_prefix: "/vineyard"		Optional: {}

VineyardCompTemplateSpec

VineyardCompTemplateSpec is the common configurations for vineyard components including Master and Worker.

Appears in:

• MasterSpec
• VineyardRuntimeSpec

Field	Description	Default	Validation
replicas integer	The replicas of Vineyard component. If not specified, defaults to 1. For worker, the replicas should not be greater than the number of nodes in the cluster		Minimum: 1 Optional: {}
image string	The image of Vineyard component. For Master, the default image is registry.aliyuncs.com/vineyard/vineyardd For Worker, the default image is registry.aliyuncs.com/vineyard/vineyardd The default container registry is docker.io, you can change it by setting the image field		Optional: {}
imageTag string	The image tag of Vineyard component. For Master, the default image tag is v0.22.2. For Worker, the default image tag is v0.22.2.		Optional: {}
imagePullPolicy string	The image pull policy of Vineyard component. Default is IfNotPresent.		Optional: {}
nodeSelector object (keys:string, values:string)	NodeSelector is a selector to choose which nodes to launch the Vineyard component. E,g. {"disktype": "ssd"}		Optional: {}
ports object (keys:string, values:integer)	Ports used by Vineyard component. For Master, the default client port is 2379 and peer port is 2380. For Worker, the default rpc port is 9600 and the default exporter port is 9144.		Optional: {}
env object (keys:string, values:string)	Environment variables that will be used by Vineyard component. For Master, refer to Etcd Configuration for more info Default is not set.		Optional: {}
options object (keys:string, values:string)	Configurable options for Vineyard component. For Master, there is no configurable options. For Worker, support the following options. vineyardd.reserve.memory: (Bool) where to reserve memory for vineyardd If set to true, the memory quota will be counted to the vineyardd rather than the application. etcd.prefix: (String) the prefix of etcd key for vineyard objects wait.etcd.timeout: (String) the timeout period before waiting the etcd to be ready, in seconds Default value is as follows. vineyardd.reserve.memory: "true" etcd.prefix: "/vineyard" wait.etcd.timeout: "120"		Optional: {}
resources ResourceRequirements	Resources contains the resource requirements and limits for the Vineyard component. Default is not set. For Worker, when the options contains vineyardd.reserve.memory=true, the resources.request.memory for worker should be greater than tieredstore.levels[0].quota(aka vineyardd shared memory)		Optional: {}
volumeMounts VolumeMount array	VolumeMounts specifies the volumes listed in ".spec.volumes" to mount into the vineyard runtime component's filesystem. It is useful for specifying a persistent storage. Default is not set.		Optional: {}
podMetadata PodMetadata	PodMetadata defines labels and annotations that will be propagated to Vineyard's pods including Master and Worker. Default is not set.		Optional: {}
networkMode NetworkMode	Whether to use hostnetwork or not Default is HostNetwork		Enum: [HostNetwork ContainerNetwork] Optional: {}

VineyardRuntime

VineyardRuntime is the Schema for the VineyardRuntimes API

Field	Description	Default	Validation
apiVersion string	data.fluid.io/v1alpha1
kind string	VineyardRuntime
metadata ObjectMeta	Refer to Kubernetes API documentation for fields of metadata.
spec VineyardRuntimeSpec

VineyardRuntimeSpec

VineyardRuntimeSpec defines the desired state of VineyardRuntime

Appears in:

• VineyardRuntime

Field	Description	Default	Validation
master MasterSpec	Master holds the configurations for Vineyard Master component Represents the Etcd component in Vineyard		Optional: {}
worker VineyardCompTemplateSpec	Worker holds the configurations for Vineyard Worker component Represents the Vineyardd component in Vineyard		Optional: {}
replicas integer	The replicas of the worker, need to be specified If worker.replicas and the field are both specified, the field will be respected
fuse VineyardClientSocketSpec	Fuse holds the configurations for Vineyard client socket. Note that the "Fuse" here is kept just for API consistency, VineyardRuntime mount a socket file instead of a FUSE filesystem to make data cache available. Applications can connect to the vineyard runtime components through IPC or RPC. IPC is the default way to connect to vineyard runtime components, which is more efficient than RPC. If the socket file is not mounted, the connection will fall back to RPC.		Optional: {}
tieredstore TieredStore	Tiered storage used by vineyardd The MediumType can only be MEM and SSD MEM actually represents the shared memory of vineyardd. SSD represents the external storage of vineyardd. Default is as follows. tieredstore: levels: - level: 0 mediumtype: MEM quota: 4Gi Choose hostpath as the external storage of vineyardd. tieredstore: levels: - level: 0 mediumtype: MEM quota: 4Gi high: "0.8" low: "0.3" - level: 1 mediumtype: SSD quota: 10Gi volumeType: Hostpath path: /var/spill-path		Optional: {}
disablePrometheus boolean	Disable monitoring metrics for Vineyard Runtime Default is false		Optional: {}
podMetadata PodMetadata	PodMetadata defines labels and annotations that will be propagated to Vineyard's pods.		Optional: {}
volumes Volume array	Volumes is the list of Kubernetes volumes that can be mounted by the vineyard components (Master and Worker). Default is null.		Optional: {}

VolumeMediumSource

VolumeMediumSource describes a Kubernetes volume as a storage medium. Only one of its members may be specified.

Appears in:

• MediumSource

Field	Description	Default	Validation
hostPath HostPathVolumeSource	HostPath represents a pre-existing file or directory on the host machine that is directly exposed to the container. More info: https://kubernetes.io/docs/concepts/storage/volumes#hostpath		Optional: {}
emptyDir EmptyDirVolumeSource	EmptyDir represents a temporary directory that shares a pod's lifetime. More info: https://kubernetes.io/docs/concepts/storage/volumes#emptydir		Optional: {}
ephemeral EphemeralVolumeSource	Ephemeral represents a volume that is handled by a cluster storage driver. The volume's lifecycle is tied to the pod that defines it. More info: https://kubernetes.io/docs/concepts/storage/ephemeral-volumes/		Optional: {}

WaitingStatus

Appears in:

• OperationStatus

Field	Description	Default	Validation
operationComplete boolean	OperationComplete indicates if the preceding operation is complete