MongoDB Database Resource Specification

Required

apiVersion

Type: string

Version of the MongoDB resource schema.

kind

Type: string

Kind of MongoDB resource to create. Set this to MongoDB

metadata.name

Type: string

Name of the MongoDB resource that you create.

Resource names must be 44 characters or less.

spec.credentials

Type: string

Required. Name of the Kubernetes secret you created as Ops Manager API authentication credentials for the Kubernetes Operator to communicate with Cloud Manager or Ops Manager.

The Ops Manager Kubernetes Secret object holding the Credentials must exist on the same Namespace as the resource you want to create.

Important

Operator manages changes to the Secret

The Kubernetes Operator tracks any changes to the Secret and reconciles the state of the MongoDB resource.

spec.persistent

Type: boolean

Default: True

Warning

Grant your containers permission to write to your Persistent Volume. The Kubernetes Operator sets fsGroup = 2000, runAsUser = 2000, and runAsNonRoot = true in securityContext. Kubernetes Operator sets fsgroup equal to runAsUser to make the volume writable for a user that runs the main process in the container. To learn more, see Configure a Security Context for a Pod or Container and the related discussion in the Kubernetes documentation. If redeploying the resource doesn't fix issues with your Persistent Volume, contact MongoDB Support.

Note

If you do not use Persistent Volumes, the Disk Usage and Disk IOPS charts cannot be displayed in either the Processes tab on the Deployment page or in the Metrics page when reviewing the data for this deployment.

spec.type

Type: string

Type of MongoDB resource to create. Accepted values are:

• Standalone

• ReplicaSet

• ShardedCluster

spec.version

Type: string

Version of MongoDB that you installed on this MongoDB resource.

Important

Ensure that you choose a compatible MongoDB Server version.

Compatible versions differ depending on the base image that the MongoDB database resource uses.

Note

If you update this value to a later version of MongoDB for your database resources, the Feature Compatibility Version (FCV) changes automatically to this version unless you set spec.featureCompatibilityVersion to the original version. Consider setting spec.featureCompatibilityVersion to the original version to give yourself the option to downgrade if necessary.

Conditional

Every resource must use one of the following settings:

spec.opsManager.configMapRef.name

Type: string

Name of the ConfigMap with the Cloud Manager or Ops Manager connection configuration. The spec.cloudManager.configMapRef.name setting is an alias for this setting and can be used in its place.

Note

This value must exist on the same namespace as the resource you want to create.

Important

Operator manages changes to the ConfigMap

The Kubernetes Operator tracks any changes to the ConfigMap and reconciles the state of the MongoDB resource.

spec.cloudManager.configMapRef.name

Type: string

Alias for spec.opsManager.configMapRef.name.

Optional

Every resource type may use the following settings:

spec.featureCompatibilityVersion

Type: string

Limits changes to data that occur with an upgrade to a new major version. This allows you to downgrade to the previous major version. To learn more about feature compatibility, see setFeatureCompatibilityVersion in the MongoDB Manual.

spec.clusterDomain

Type: string

Default: cluster.local

Domain name of the Kubernetes cluster where you deploy the Kubernetes Operator. When Kubernetes creates a StatefulSet, the Kubernetes assigns each Pod a FQDN. To update Cloud Manager or Ops Manager, the Kubernetes Operator calculates the FQDN for each Pod using a provided cluster name. Kubernetes doesn't provide an API to query these hostnames.

Warning

You must set spec.clusterDomain if your Kubernetes cluster has a default domain other than the default cluster.local. If you neither use the default nor set the spec.clusterDomain option, the Kubernetes Operator might not function as expected.

spec.clusterName

Type: string

Default: cluster.local

Important

spec.clusterName is Deprecated

Use spec.clusterDomain instead.

Domain name of the Kubernetes cluster where you deploy the Kubernetes Operator. When Kubernetes creates a StatefulSet, the Kubernetes assigns each Pod a FQDN. To update Cloud Manager or Ops Manager, the Kubernetes Operator calculates the FQDN for each Pod using a provided cluster name. Kubernetes doesn't provide an API to query these hostnames.

Warning

You must set spec.clusterDomain if your Kubernetes cluster has a default domain other than the default cluster.local. If you neither use the default nor set the spec.clusterDomain option, the Kubernetes Operator might not function as expected.

metadata.namespace

Type: string

Kubernetes namespace where you create this MongoDB resource and other objects.

spec.service

Type: string

Default: <resource_name>+"-svc" and <resource_name>+"-svc-external"

Important

spec.service is Deprecated

Use spec.statefulSet.spec.serviceName instead.

Name of the Kubernetes service to be created or used for a StatefulSet. If the service with this name already exists, the MongoDB Enterprise Kubernetes Operator does not delete or recreate it. This setting lets you create your own custom services and lets the Kubernetes Operator reuse them.

spec.logLevel

Type: string

Default: INFO

Configures the level of Automation Agent logging inside the Pod. Accepted values include:

• DEBUG

• INFO

• WARN

• ERROR

• FATAL

spec.security.authentication.ignoreUnknownUsers

Type: boolean

Default: false

Determines whether you can modify database users that were not configured through the Kubernetes Operator, or the Cloud Manager or Ops Manager user interface.

To manage database users directly through the mongod or mongos, set this setting to true.

Standalone Settings

spec.additionalMongodConfig

Type: collection

Additional configuration options with which you want to start MongoDB processes.

The Kubernetes Operator supports all configuration options that the MongoDB version you deploy through the MongoDB Agent supports, except that the Kubernetes Operator overrides values that you provide for any of the following options:

• net.tls.certificateKeyFile

• net.tls.clusterFile

• net.tls.PEMKeyFile

• replication.replSetName

• security.clusterAuthMode

• sharding.clusterRole

• storage.dbPath

• systemLog.destination

• systemLog.path

To learn more about the configuration options that the Kubernetes Operator owns, see MongoDB Kubernetes Operator Exclusive Settings.

To learn which configuration options you can use, see Advanced Options for MongoDB Deployments in the Ops Manager documentation.

spec.agent

Type: collection

MongoDB Agent configuration settings for MongoDB database resource.

spec.agent.startupOptions

Type: collection

MongoDB Agent settings with which you want to start MongoDB database resource.

You must provide MongoDB Agent settings as key-value pairs. The values must be strings.

For a list of supported MongoDB Agent settings, see:

• MongoDB Agent Settings for Cloud Manager projects.

• MongoDB Agent Settings for the Ops Manager version you deployed with the Kubernetes Operator.

1
---
2
apiVersion: mongodb.com/v1
3
kind: MongoDB
4
metadata:
5
  name: my-standalone
6
spec:
7
  version: "4.4.0-ent"
8
  service: my-service
9
10
  opsManager:
11
    configMapRef:
12
      name: my-project
13
  credentials: my-credentials
14
  type: Standalone
15
16
  persistent: true
17
  agent:
18
    startupOptions:
19
      maxLogFiles: "30"
20
      dialTimeoutSeconds: "40"
21
...

spec.exposedExternally

Type: boolean

Default: false

Important

spec.exposedExternally is deprecated starting in Kubernetes Operator version 1.19 and will be removed in a future Kubernetes Operator release. To configure how your deployment should be exposed externally, use the spec.externalAccess settings.

Determines whether a multi-Kubernetes-cluster deployment is exposed outside the Kubernetes clusters. Setting spec.exposedExternally: true is equivalent to setting spec.externalAccess: {}. See spec.externalAccess for details.

spec.externalAccess

Type: collection

Specification to expose your cluster for external connections. To learn how to connect to your MongoDB resource from outside of the Kubernetes cluster, see Connect to a MongoDB Database Resource from Outside Kubernetes.

If you add spec.externalAccess, the Kubernetes Operator creates an external service for each Pod in a replica set. External services provide an external entry point for each MongoDB database Pod in a cluster. Each external service has selectors that match the external service to a specific Pod.

If you add this setting without any values, the Kubernetes Operator creates an external service with the following default values:

Field	Value	Description
Name	<pod-name>-svc-external	Name of the external service. You can't change this value.
Type	LoadBalancer	Creates an external LoadBalancer service.
Port	<Port Number>	A port for mongod.
publishNotReadyAddress	true	Specifies that DNS records are created even if the Pod isn't ready. Do not set to false for any database Pod.

Note

If you set spec.externalAccess.externalDomain, the external service adds another port (Port Number + 1) for backups.

spec.externalAccess.externalService

Type: collection

Specification for overriding the default values in spec.externalAccess.

When you set the spec.externalAccess setting, the Kubernetes Operator automatically creates an external load balancer service with default values. You can override certain values or add new values depending on your needs. For example, if you intend to create NodePort services and don't need a load balancer, you must configure overrides in your Kubernetes specification:

externalAccess:
  externalService:
    annotations:
      # cloud-specific annotations for the service
    spec:
      type: NodePort # default is LoadBalancer
      # you can specify other spec overrides if necessary

For more information about the Kubernetes specification, see ServiceSpec in the Kubernetes documentation.

spec.externalAccess.externalService.annotations

Type: collection

Key-value pairs that let you add cloud provider-specific configuration settings to all clusters in your deployment.

To learn more about annotations, see the Kubernetes documentation and the documentation for your Kubernetes cloud provider.

spec.externalAccess.externalService.spec

Type: collection

Configuration for the ServiceSpec. To learn more, see spec.externalAccess.externalService.

spec.podSpec.persistence.single

Type: collection

Has Kubernetes Operator create one Persistent Volume Claim and mount all three directories for data, journal, and logs to the same Persistent Volume.

Note

• You must set the values in this collection if spec.persistent : true.

• You may set this collection or the persistence.multiple collections but not both.

Scalar	Data Type	Description
labelSelector	string	Tag used to bind mounted volumes to directories.
storage	string	Minimum size of Persistent Volume that should be mounted. This value is expressed as an integer followed by a unit of storage in JEDEC notation. Default value is 16Gi. Example If standalone deployment in requires 60 gigabytes of storage space, set this value to 60Gi.
storageClass	string	Type of storage specified in a Persistent Volume Claim. You may create this storage type as a StorageClass object before using it in this object specification. Note Make sure to set the StorageClass reclaimPolicy to Retain. This ensures that data is retained when a Persistent Volume Claim is removed.

spec.podSpec.persistence.multiple.data

Type: collection

Has Kubernetes Operator create a Persistent Volume Claim and mount a directory for data to its own Persistent Volume.

Note

• You must set the values in this collection if spec.persistent : true.

• You may set this collection or the persistence.single collection but not both.

Scalar	Data Type	Description
labelSelector	string	Tag used to bind mounted volumes to directories.
storage	string	Minimum storage capacity that must be available on a Kubernetes node to host standalone deployment on Kubernetes. This value is expressed as an integer followed by a unit of storage in JEDEC notation. Default value is 16Gi. Example If this MongoDB resource requires 60 gigabytes of storage space, set this value to 60Gi.
storageClass	string	Type of storage needed for standalone deployment. You may create this storage type as a StorageClass object before using it in this object specification. Note Make sure to set the StorageClass reclaimPolicy to Retain. This ensures that data is retained when a Persistent Volume Claim is removed.

spec.podSpec.persistence.multiple.journal

Type: collection

Has Kubernetes Operator create a Persistent Volume Claim and mount a directory for journal to its own Persistent Volume.

Note

• You must set the values in this collection if spec.persistent : true.

• You may set this collection or the persistence.single collection but not both.

Scalar	Data Type	Description
labelSelector	string	Tag used to bind mounted volumes to directories.
storage	string	Minimum storage capacity that must be available on a Kubernetes node to host standalone deployment on Kubernetes. This value is expressed as an integer followed by a unit of storage in JEDEC notation. Default value is 1Gi. Example If this MongoDB resource requires 60 gigabytes of storage space, set this value to 60Gi.
storageClass	string	Type of storage needed for standalone deployment. You may create this storage type as a StorageClass object before using it in this object specification. Note Make sure to set the StorageClass reclaimPolicy to Retain. This ensures that data is retained when a Persistent Volume Claim is removed.

spec.podSpec.persistence.multiple.logs

Type: collection

Has Kubernetes Operator create a Persistent Volume Claim and mount a directory for logs to its own Persistent Volume.

Note

• You must set the values in this collection if spec.persistent : true.

• You may set this collection or the persistence.single collection but not both.

Scalar	Data Type	Description
labelSelector	string	Tag used to bind mounted volumes to directories.
storage	string	Minimum storage capacity that must be available on a Kubernetes node to host standalone deployment on Kubernetes. This value is expressed as an integer followed by a unit of storage in JEDEC notation. Default value is 3Gi. Example If this MongoDB resource requires 60 gigabytes of storage space, set this value to 60Gi.
storageClass	string	Type of storage needed for standalone deployment. You may create this storage type as a StorageClass object before using it in this object specification. Note Make sure to set the StorageClass reclaimPolicy to Retain. This ensures that data is retained when a Persistent Volume Claim is removed.

spec.podSpec.podTemplate.affinity.nodeAffinity

Type: Struct

Kubernetes rule to place Pods for replica set on a specific range of nodes.

For optimized read-write performance, use node affinity rules that restrict Pods to run on particular nodes, or to prefer to run on particular nodes.

spec.podSpec.podTemplate.affinity.podAffinity

Type: Struct

Kubernetes rule to determine whether multiple MongoDB resource Pods must be co-located with other Pods. To learn more about the use cases, see Affinity and Anti-Affinity in the Kubernetes documentation.

spec.podSpec.podTemplate.affinity.podAntiAffinity

Type: Struct

Default: kubernetes.io/hostname

Sets a rule to spread Pods hosting MongoDB resource to different locations. A location can be a single node, rack, or region. By default, Kubernetes Operator tries to spread pods across different nodes.

spec.podSpec.podTemplate.affinity.podAntiAffinity.requiredDuringSchedulingIgnoredDuringExecution.topologyKey

Type: Struct

Default: kubernetes.io/hostname

This key defines which label is used to determine which topology domain a node belongs to.

spec.podSpec.podTemplate

Type: collection

Template for the Kubernetes Pods that the MongoDB Enterprise Kubernetes Operator creates for MongoDB database resources.

Template values take precedence over values specified in spec.podSpec.

Note

The Kubernetes Operator doesn't validate the fields you provide in spec.podSpec.podTemplate.

spec.podSpec.podTemplate.metadata

Type: collection

Metadata for the Kubernetes Pods that the MongoDB Enterprise Kubernetes Operator creates for MongoDB database resources.

To review which fields you can add to spec.podSpec.podTemplate.metadata, see the Kubernetes documentation.

spec.podSpec.podTemplate.spec

Type: collection

Specifications of the Kubernetes Pods that the MongoDB Enterprise Kubernetes Operator creates for MongoDB database resources.

To review which fields you can add to spec.podSpec.podTemplate.spec, see the Kubernetes PodSpec v1 core API.

Note

When you add containers to spec.podSpec.podTemplate.spec.containers, the Kubernetes Operator adds them to the Kubernetes pod. These containers are appended to MongoDB database resources containers in the pod.

Use this setting to specify the CPU and RAM allocations for each pod. For examples, see the samples on GitHub.

Replica Set Settings

The following settings apply to replica set resource types:

spec.backup

Type: collection

The collection container for spec.backup.mode, which enables continuous backups for MongoDB resources in Kubernetes Operator.

spec.backup.assignmentLabels

Type: array

A comma-separated list of labels to assign backup daemons, oplog stores, blockstores, S3 snapshot stores, and file system stores to specific projects or groups. Use assignment labels to identify that specific backup stores are associated with particular projects.

If you set assignment labels using the Kubernetes Operator, the values that you set in the Kubernetes configuration file for assignment labels override the values defined in the Ops Manager UI. Assignment labels that you don't set using the Kubernetes Operator continue to use the values set in the Ops Manager UI.

spec.backup.mode

Type: string

Enables continuous backups for a MongoDB resource. Possible values are enabled, disabled, and terminated.

Note

The spec.backup.mode setting relies on Backup that is enabled in the Ops Manager and requires that spec.backup.enabled value in the Ops Manager resource specification is set to true.

After you enable continuous backups for your MongoDB resource with spec.backup.mode, you can check the backup status.

spec.backup.encryption

Type: object

Object that contains the backup encryption configuration settings.

spec.backup.encryption.kmip

Type: object

Object that contains the KMIP backup encryption configuration settings. To learn more, see Configure KMIP Backup Encryption for Ops Manager.

spec.backup.encryption.kmip.client

Type: object

Object that contains the KMIP backup encryption client configuration settings.

spec.backup.encryption.kmip.client.clientCertificatePrefix

Type: string

spec.backup.snapshotSchedule

Type: collection

Collection container for snapshot schedule settings for continuous backups for MongoDB resources in Kubernetes Operator.

spec.backup.snapshotSchedule.snapshotIntervalHours

Type: number

Number of hours between snapshots. You can set a value of 6, 8, 12, or 24.

spec.backup.snapshotSchedule.snapshotRetentionDays

Type: number

Number of days to keep recent snapshots. You can set a value between 2 and 5, inclusive.

spec.backup.snapshotSchedule.dailySnapshotRetentionDays

Type: number

Number of days to keep daily snapshots. You can set a value between 1 and 365, inclusive. Setting the value to 0 disables this rule.

spec.backup.snapshotSchedule.weeklySnapshotRetentionWeeks

Type: number

Number of weeks to keep weekly snapshots. You can set a value between 1 and 52, inclusive. Setting the value to 0 disables this rule.

spec.backup.snapshotSchedule.monthlySnapshotRetentionMonths

Type: number

Number of months to keep monthly snapshots. You can set a value between 1 and 36, inclusive. Setting the value to 0 disables this rule.

spec.backup.snapshotSchedule.pointInTimeWindowHours

Type: number

Number of hours in the past for which you can create a point-in-time snapshot.

spec.backup.snapshotSchedule.referenceHourOfDay

Type: number

UTC hour of the day to schedule snapshots using a 24 hour clock. You can set a value between 0 and 23, inclusive.

spec.backup.snapshotSchedule.referenceMinuteOfHour

Type: number

UTC minute of the hour to schedule snapshots. You can set a value between 0 and 59, inclusive.

spec.backup.snapshotSchedule.fullIncrementalDayOfWeek

Type: string

Day of the week when Ops Manager takes a full snapshot. This setting ensures a recent complete backup. Ops Manager sets the default value to SUNDAY.

spec.clusterName

Type: string

Default: cluster.local

Important

spec.clusterName is Deprecated

Use spec.clusterDomain instead.

Domain name of the Kubernetes cluster where you deploy the Kubernetes Operator. When Kubernetes creates a StatefulSet, the Kubernetes assigns each Pod a FQDN. To update Cloud Manager or Ops Manager, the Kubernetes Operator calculates the FQDN for each Pod using a provided cluster name. Kubernetes doesn't provide an API to query these hostnames.

Warning

You must set spec.clusterDomain if your Kubernetes cluster has a default domain other than the default cluster.local. If you neither use the default nor set the spec.clusterDomain option, the Kubernetes Operator might not function as expected.

spec.connectivity.replicaSetHorizons

Type: collection

Allows you to provide different DNS settings for client applications and the MongoDB Agents. The Kubernetes Operator uses split horizon DNS for replica set members. This feature allows communication both within the Kubernetes cluster and from outside Kubernetes.

You may add multiple external mappings per host.

Note

Split Horizon Requirements

• Make sure that each value in this array is unique.

• Make sure that the number of entries in this array matches the value given in spec.members.

• Provide a value for the spec.security.certsSecretPrefix setting to enable TLS. This method to use split horizons requires the Server Name Indication extension of the TLS protocol.

• Configure the routing for external hostnames.

Example

In this example, the replica set members communicate amongst themselves on the example-localhost horizon. Clients communicate with the replica set using the example-website horizon.

1
---
2
apiVersion: mongodb.com/v1
3
kind: MongoDB
4
metadata:
5
  name: <my-replica-set>
6
spec:
7
  members: 3
8
  version: "4.2.2-ent"
9
  type: ReplicaSet
10
  opsManager:
11
    configMapRef:
12
      name: <configMap.metadata.name>
13
  credentials: <mycredentials>
14
  persistent: true

15
  security:
16
    tls:
17
      enabled: true
18
  connectivity:
19
    replicaSetHorizons:
20
      - "example-website": "web1.example.com:30907"
21
      - "example-website": "web2.example.com:32350"
22
      - "example-website": "web3.example.com:31185"
23
...

spec.externalAccess.externalDomain

Type: string

An external domain used to externally expose your replica set deployment.

By default, each replica set member uses the Kubernetes Pod's FQDN (*.svc.cluster.local) as the default hostname. However, if you add an external domain to this setting, the replica set uses a hostname that is a subdomain of the specified domain instead. This hostname uses the following format:

<replica-set-name>-<pod-idx>.<externalDomain>

For example:

replica-set-1.example.com

After you deploy the replica set with this setting, the Kubernetes Operator uses the hostname with the external domain to override the processes[n].hostname field in the Ops Manager automation configuration. Then, the MongoDB Agent uses this hostname to connect to mongod.

To specify other hostnames for connecting to the replica set, you can use the spec.connectivity.replicaSetHorizons setting. However, the following connections still use the hostname with the external domain:

• The MongoDB Agent to connect to mongod.

• mongod to connect to other mongod instances.

Warning

Specifying this field changes how Ops Manager registers mongod processes. You can specify this field only for new replica set deployments starting in Kubernetes Operator version 1.19. You can't change the value of this field or any processes[n].hostname fields in the Ops Manager automation configuration for a running replica set deployment.

spec.members

Type: integer

Required. Number of Members of the Replica Set.

spec.memberConfig

Type: collection

Specification for each MongoDB replica set member deployed from the MongoDB resource.

The order of the elements in the array must reflect the order of members in the replica set. For example, the first element of the array affects the Pod at index 0, the second element affects index 1, and so on.

Example

Consider the following example specification for a three-member replica set:

spec:
  memberConfig:
    - votes: 1
      priority: "0.5"
      tags:
        tag1: "value1"
        environment: "prod"
    - votes: 1
      priority: "1.5"
      tags:
        tag2: "value2"
        environment: "prod"
    - votes: 0
      priority: "0.5"
      tags:
        tag2: "value2"
        environment: "prod"

spec.memberConfig.priority

Type: string

Number that indicates the relative eligibility of a MongoDB replica set member to become a primary.

Specify a higher value to make a member more eligible to become primary, and lower values to make the member less eligible. For example, a member with a spec.clusterSpecList.memberConfig.priority of "1.5" is more likely than "0.5" to become a primary. A priority of "0" is ineligible to become primary. For more details, see Member Priority.

spec.memberConfig.tags

Type: map

Map of replica set tags for directing read and write operations to specific members of your MongoDB replica set.

spec.memberConfig.votes

Type: number

Determines whether a MongoDB replica set member can vote in an election. Set to 1 to allow the member to vote. Set to 0 to exclude the member from an election.

The following settings apply only to replica set resource types:

spec.backup.autoTerminateOnDeletion

Type: boolean

Flag that indicates whether the Kubernetes Operator stops and terminates the backup when you delete a MongoDB resource. If omitted, the default value is false. Setting this flag to true is useful when you want to delete the MongoDB custom resource while the spec.backup.mode setting is set to enabled.

Sharded Cluster Settings

The following settings apply only to sharded cluster resource types:

spec.backup.snapshotSchedule.clusterCheckpointIntervalMin

Type: number

Number of minutes between successive cluster checkpoints. This setting applies only to sharded clusters that run MongoDB with FCV of 4.0 or earlier. This number determines the granularity of point-in-time restores for sharded clusters. You can set a value of 15, 30, or 60.

spec.configServerCount

Type: integer

Required. Number of members in the config server.

spec.configSrv.additionalMongodConfig

Type: collection

Additional configuration options with which you want to start each config server member.

The Kubernetes Operator supports all configuration options that the MongoDB version you deploy through the MongoDB Agent supports, except that the Kubernetes Operator overrides values that you provide for any of the following options:

• net.tls.certificateKeyFile

• net.tls.clusterFile

• net.tls.PEMKeyFile

• replication.replSetName

• security.clusterAuthMode

• sharding.clusterRole

• storage.dbPath

• systemLog.destination

• systemLog.path

To learn more about the configuration options that the Kubernetes Operator owns, see MongoDB Kubernetes Operator Exclusive Settings.

To learn which configuration options you can use, see Advanced Options for MongoDB Deployments in the Ops Manager documentation.

spec.configSrv.agent

Type: collection

MongoDB Agent configuration settings for each config server member.

spec.configSrv.agent.startupOptions

Type: collection

MongoDB Agent settings with which you want to start each config server member.

You must provide MongoDB Agent settings as key-value pairs. The values must be strings.

For a list of supported MongoDB Agent settings, see:

• MongoDB Agent Settings for Cloud Manager projects.

• MongoDB Agent Settings for the Ops Manager version you deployed with the Kubernetes Operator.

1
---
2
apiVersion: mongodb.com/v1
3
kind: MongoDB
4
metadata:
5
  name: my-sharded-cluster-options
6
spec:
7
  version: "4.4.0-ent"
8
  type: ShardedCluster
9
  opsManager:
10
    configMapRef:
11
      name: my-project
12
  credentials: my-credentials
13
  persistent: true
14
  shardCount: 2
15
  mongodsPerShardCount: 3
16
  mongosCount: 2
17
  configServerCount: 1
18
19
  mongos:
20
    agent:
21
      startupOptions:
22
        maxLogFiles: "30"
23
24
  configSrv:
25
     agent:
26
       startupOptions:
27
         dialTimeoutSeconds: "40"
28
  shard:
29
     agent:
30
       startupOptions:
31
         serverSelectionTimeoutSeconds: "20"
32
...

spec.configSrvPodSpec.persistence.single

Type: collection

Has Kubernetes Operator create one Persistent Volume Claim and mount all three directories for data, journal, and logs to the same Persistent Volume.

Note

• You must set the values in this collection if spec.persistent : true.

• You may set this collection or the persistence.multiple collections but not both.

Scalar	Data Type	Description
labelSelector	string	Tag used to bind mounted volumes to directories.
storage	string	Minimum size of Persistent Volume that should be mounted. This value is expressed as an integer followed by a unit of storage in JEDEC notation. Default value is 5Gi. Example If each config server member in requires 60 gigabytes of storage space, set this value to 60Gi.
storageClass	string	Type of storage specified in a Persistent Volume Claim. You may create this storage type as a StorageClass object before using it in this object specification. Note Make sure to set the StorageClass reclaimPolicy to Retain. This ensures that data is retained when a Persistent Volume Claim is removed.

spec.configSrvPodSpec.persistence.multiple.data

Type: collection

Has Kubernetes Operator create a Persistent Volume Claim and mount a directory for data to its own Persistent Volume.

Note

• You must set the values in this collection if spec.persistent : true.

• You may set this collection or the persistence.single collection but not both.

Scalar	Data Type	Description
labelSelector	string	Tag used to bind mounted volumes to directories.
storage	string	Minimum storage capacity that must be available on a Kubernetes node to host each config server member on Kubernetes. This value is expressed as an integer followed by a unit of storage in JEDEC notation. Default value is 16Gi. Example If this MongoDB resource requires 60 gigabytes of storage space, set this value to 60Gi.
storageClass	string	Type of storage needed for each config server member. You may create this storage type as a StorageClass object before using it in this object specification. Note Make sure to set the StorageClass reclaimPolicy to Retain. This ensures that data is retained when a Persistent Volume Claim is removed.

spec.configSrvPodSpec.persistence.multiple.journal

Type: collection

Has Kubernetes Operator create a Persistent Volume Claim and mount a directory for journal to its own Persistent Volume.

Note

• You must set the values in this collection if spec.persistent : true.

• You may set this collection or the persistence.single collection but not both.

Scalar	Data Type	Description
labelSelector	string	Tag used to bind mounted volumes to directories.
storage	string	Minimum storage capacity that must be available on a Kubernetes node to host each config server member on Kubernetes. This value is expressed as an integer followed by a unit of storage in JEDEC notation. Default value is 1Gi. Example If this MongoDB resource requires 60 gigabytes of storage space, set this value to 60Gi.
storageClass	string	Type of storage needed for each config server member. You may create this storage type as a StorageClass object before using it in this object specification. Note Make sure to set the StorageClass reclaimPolicy to Retain. This ensures that data is retained when a Persistent Volume Claim is removed.

spec.configSrvPodSpec.persistence.multiple.logs

Type: collection

Has Kubernetes Operator create a Persistent Volume Claim and mount a directory for logs to its own Persistent Volume.

Note

• You must set the values in this collection if spec.persistent : true.

• You may set this collection or the persistence.single collection but not both.

Scalar	Data Type	Description
labelSelector	string	Tag used to bind mounted volumes to directories.
storage	string	Minimum storage capacity that must be available on a Kubernetes node to host each config server member on Kubernetes. This value is expressed as an integer followed by a unit of storage in JEDEC notation. Default value is 3Gi. Example If this MongoDB resource requires 60 gigabytes of storage space, set this value to 60Gi.
storageClass	string	Type of storage needed for each config server member. You may create this storage type as a StorageClass object before using it in this object specification. Note Make sure to set the StorageClass reclaimPolicy to Retain. This ensures that data is retained when a Persistent Volume Claim is removed.

spec.configSrvPodSpec.podTemplate

Type: collection

Template for the Kubernetes Pods that the MongoDB Enterprise Kubernetes Operator creates for each config server member.

Template values take precedence over values specified in spec.configSrvPodSpec.

Note

The Kubernetes Operator doesn't validate the fields you provide in spec.configSrvPodSpec.podTemplate.

spec.configSrvPodSpec.podTemplate.affinity.podAffinity

Type: collection

Kubernetes rule to determine whether multiple MongoDB resource Pods must be co-located with other Pods. To learn more about the use cases, see Affinity and Anti-Affinity in the Kubernetes documentation.

spec.configSrvPodSpec.podTemplate.affinity.nodeAffinity

Type: collection

Kubernetes rule to place Pods for replica set on a specific range of nodes.

For optimized read-write performance, use node affinity rules that restrict Pods to run on particular nodes, or to prefer to run on particular nodes.

spec.configSrvPodSpec.podTemplate.affinity.podAntiAffinity

Type: string

Default: kubernetes.io/hostname

Sets a rule to spread Pods hosting MongoDB resource to different locations. A location can be a single node, rack, or region. By default, Kubernetes Operator tries to spread pods across different nodes.

spec.configSrvPodSpec.podTemplate.affinity.podAntiAffinity.requiredDuringSchedulingIgnoredDuringExecution.topologyKey

Type: string

Default: kubernetes.io/hostname

This key defines which label is used to determine which topology domain a node belongs to.

spec.configSrvPodSpec.podTemplate.metadata

Type: collection

Metadata for the Kubernetes Pods that the MongoDB Enterprise Kubernetes Operator creates for each config server member.

To review which fields you can add to spec.configSrvPodSpec.podTemplate.metadata, see the Kubernetes documentation.

spec.configSrvPodSpec.podTemplate.spec

Type: collection

Specifications of the Kubernetes Pods that the MongoDB Enterprise Kubernetes Operator creates for each config server member.

To review which fields you can add to spec.configSrvPodSpec.podTemplate.spec, see the Kubernetes PodSpec v1 core API.

Note

When you add containers to spec.configSrvPodSpec.podTemplate.spec.containers, the Kubernetes Operator adds them to the Kubernetes pod. These containers are appended to each config server member containers in the pod.

Use this setting to specify the CPU and RAM allocations for each pod. For examples, see the samples on GitHub.

spec.mongodsPerShardCount

Type: integer

Required. Number of members per shard.

spec.mongosCount

Type: integer

Required. Number of mongos instances in the sharded cluster.

spec.mongos.additionalMongodConfig

Type: collection

Additional configuration options with which you want to start each mongos instance.

The Kubernetes Operator supports all configuration options that the MongoDB version you deploy through the MongoDB Agent supports, except that the Kubernetes Operator overrides values that you provide for any of the following options:

• net.tls.certificateKeyFile

• net.tls.clusterFile

• net.tls.PEMKeyFile

• replication.replSetName

• security.clusterAuthMode

• sharding.clusterRole

• storage.dbPath

• systemLog.destination

• systemLog.path

To learn more about the configuration options that the Kubernetes Operator owns, see MongoDB Kubernetes Operator Exclusive Settings.

To learn which configuration options you can use, see Advanced Options for MongoDB Deployments in the Ops Manager documentation.

spec.mongos.agent

Type: collection

MongoDB Agent configuration settings for each mongos instance.

spec.mongos.agent.startupOptions

Type: collection

MongoDB Agent settings with which you want to start each mongos instance.

You must provide MongoDB Agent settings as key-value pairs. The values must be strings.

For a list of supported MongoDB Agent settings, see:

• MongoDB Agent Settings for Cloud Manager projects.

• MongoDB Agent Settings for the Ops Manager version you deployed with the Kubernetes Operator.

1
---
2
apiVersion: mongodb.com/v1
3
kind: MongoDB
4
metadata:
5
  name: my-sharded-cluster-options
6
spec:
7
  version: "4.4.0-ent"
8
  type: ShardedCluster
9
  opsManager:
10
    configMapRef:
11
      name: my-project
12
  credentials: my-credentials
13
  persistent: true
14
  shardCount: 2
15
  mongodsPerShardCount: 3
16
  mongosCount: 2
17
  configServerCount: 1
18
19
  mongos:
20
    agent:
21
      startupOptions:
22
        maxLogFiles: "30"
23
24
  configSrv:
25
     agent:
26
       startupOptions:
27
         dialTimeoutSeconds: "40"
28
  shard:
29
     agent:
30
       startupOptions:
31
         serverSelectionTimeoutSeconds: "20"
32
...

spec.mongosPodSpec.podTemplate

Type: collection

Template for the Kubernetes Pods that the MongoDB Enterprise Kubernetes Operator creates for each mongos instance.

Template values take precedence over values specified in spec.mongosPodSpec.

Note

The Kubernetes Operator doesn't validate the fields you provide in spec.mongosPodSpec.podTemplate.

spec.mongosPodSpec.podTemplate.affinity.podAffinity

Type: collection

Optional. Kubernetes rule to determine if multiple MongoDB resource Pods must be co-located with other Pods.

spec.mongosPodSpec.podTemplate.affinity.nodeAffinity

Type: collection

Kubernetes rule to place Pods for replica set on a specific range of nodes.

For optimized read-write performance, use node affinity rules that restrict Pods to run on particular nodes, or to prefer to run on particular nodes.

spec.mongosPodSpec.podTemplate.affinity.podAntiAffinity

Type: string

Default: kubernetes.io/hostname

Sets a rule to spread Pods hosting MongoDB resource to different locations. A location can be a single node, rack, or region. By default, Kubernetes Operator tries to spread pods across different nodes.

spec.mongosPodSpec.podTemplate.affinity.podAntiAffinity.requiredDuringSchedulingIgnoredDuringExecution.topologyKey

Type: string

Default: kubernetes.io/hostname

This key defines which label is used to determine which topology domain a node belongs to.

spec.mongosPodSpec.podTemplate.metadata

Type: collection

Metadata for the Kubernetes Pods that the MongoDB Enterprise Kubernetes Operator creates for each mongos instance.

To review which fields you can add to spec.mongosPodSpec.podTemplate.metadata, see the Kubernetes documentation.

spec.mongosPodSpec.podTemplate.spec

Type: collection

Specifications of the Kubernetes Pods that the MongoDB Enterprise Kubernetes Operator creates for each mongos instance.

To review which fields you can add to spec.mongosPodSpec.podTemplate.spec, see the Kubernetes PodSpec v1 core API.

Note

When you add containers to spec.mongosPodSpec.podTemplate.spec.containers, the Kubernetes Operator adds them to the Kubernetes pod. These containers are appended to each mongos instance containers in the pod.

Use this setting to specify the CPU and RAM allocations for each pod. For examples, see the samples on GitHub.

spec.shardCount

Type: integer

Required. Number of shards in the sharded cluster.

spec.shard.additionalMongodConfig

Type: collection

Additional configuration options with which you want to start each sharded cluster shard member.

The Kubernetes Operator supports all configuration options that the MongoDB version you deploy through the MongoDB Agent supports, except that the Kubernetes Operator overrides values that you provide for any of the following options:

• net.tls.certificateKeyFile

• net.tls.clusterFile

• net.tls.PEMKeyFile

• replication.replSetName

• security.clusterAuthMode

• sharding.clusterRole

• storage.dbPath

• systemLog.destination

• systemLog.path

To learn more about the configuration options that the Kubernetes Operator owns, see MongoDB Kubernetes Operator Exclusive Settings.

To learn which configuration options you can use, see Advanced Options for MongoDB Deployments in the Ops Manager documentation.

spec.shard.agent

Type: collection

MongoDB Agent configuration settings for each sharded cluster shard member.

spec.shard.agent.startupOptions

Type: collection

MongoDB Agent settings with which you want to start each sharded cluster shard member.

You must provide MongoDB Agent settings as key-value pairs. The values must be strings.

For a list of supported MongoDB Agent settings, see:

• MongoDB Agent Settings for Cloud Manager projects.

• MongoDB Agent Settings for the Ops Manager version you deployed with the Kubernetes Operator.

1
---
2
apiVersion: mongodb.com/v1
3
kind: MongoDB
4
metadata:
5
  name: my-sharded-cluster-options
6
spec:
7
  version: "4.4.0-ent"
8
  type: ShardedCluster
9
  opsManager:
10
    configMapRef:
11
      name: my-project
12
  credentials: my-credentials
13
  persistent: true
14
  shardCount: 2
15
  mongodsPerShardCount: 3
16
  mongosCount: 2
17
  configServerCount: 1
18
19
  mongos:
20
    agent:
21
      startupOptions:
22
        maxLogFiles: "30"
23
24
  configSrv:
25
     agent:
26
       startupOptions:
27
         dialTimeoutSeconds: "40"
28
  shard:
29
     agent:
30
       startupOptions:
31
         serverSelectionTimeoutSeconds: "20"
32
...

spec.shardPodSpec.persistence.single

Type: collection

Has Kubernetes Operator create one Persistent Volume Claim and mount all three directories for data, journal, and logs to the same Persistent Volume.

Note

• You must set the values in this collection if spec.persistent : true.

• You may set this collection or the persistence.multiple collections but not both.

Scalar	Data Type	Description
labelSelector	string	Tag used to bind mounted volumes to directories.
storage	string	Minimum size of Persistent Volume that should be mounted. This value is expressed as an integer followed by a unit of storage in JEDEC notation. Default value is 16Gi. Example If each sharded cluster shard member in requires 60 gigabytes of storage space, set this value to 60Gi.
storageClass	string	Type of storage specified in a Persistent Volume Claim. You may create this storage type as a StorageClass object before using it in this object specification. Note Make sure to set the StorageClass reclaimPolicy to Retain. This ensures that data is retained when a Persistent Volume Claim is removed.

spec.shardPodSpec.persistence.multiple.data

Type: collection

Has Kubernetes Operator create a Persistent Volume Claim and mount a directory for data to its own Persistent Volume.

Note

• You must set the values in this collection if spec.persistent : true.

• You may set this collection or the persistence.single collection but not both.

Scalar	Data Type	Description
labelSelector	string	Tag used to bind mounted volumes to directories.
storage	string	Minimum storage capacity that must be available on a Kubernetes node to host each sharded cluster shard member on Kubernetes. This value is expressed as an integer followed by a unit of storage in JEDEC notation. Default value is 16Gi. Example If this MongoDB resource requires 60 gigabytes of storage space, set this value to 60Gi.
storageClass	string	Type of storage needed for each sharded cluster shard member. You may create this storage type as a StorageClass object before using it in this object specification. Note Make sure to set the StorageClass reclaimPolicy to Retain. This ensures that data is retained when a Persistent Volume Claim is removed.

spec.shardPodSpec.persistence.multiple.journal

Type: collection

Has Kubernetes Operator create a Persistent Volume Claim and mount a directory for journal to its own Persistent Volume.

Note

• You must set the values in this collection if spec.persistent : true.

• You may set this collection or the persistence.single collection but not both.

Scalar	Data Type	Description
labelSelector	string	Tag used to bind mounted volumes to directories.
storage	string	Minimum storage capacity that must be available on a Kubernetes node to host each sharded cluster shard member on Kubernetes. This value is expressed as an integer followed by a unit of storage in JEDEC notation. Default value is 1Gi. Example If this MongoDB resource requires 60 gigabytes of storage space, set this value to 60Gi.
storageClass	string	Type of storage needed for each sharded cluster shard member. You may create this storage type as a StorageClass object before using it in this object specification. Note Make sure to set the StorageClass reclaimPolicy to Retain. This ensures that data is retained when a Persistent Volume Claim is removed.

spec.shardPodSpec.persistence.multiple.logs

Type: collection

Has Kubernetes Operator create a Persistent Volume Claim and mount a directory for logs to its own Persistent Volume.

Note

• You must set the values in this collection if spec.persistent : true.

• You may set this collection or the persistence.single collection but not both.

Scalar	Data Type	Description
labelSelector	string	Tag used to bind mounted volumes to directories.
storage	string	Minimum storage capacity that must be available on a Kubernetes node to host each sharded cluster shard member on Kubernetes. This value is expressed as an integer followed by a unit of storage in JEDEC notation. Default value is 3Gi. Example If this MongoDB resource requires 60 gigabytes of storage space, set this value to 60Gi.
storageClass	string	Type of storage needed for each sharded cluster shard member. You may create this storage type as a StorageClass object before using it in this object specification. Note Make sure to set the StorageClass reclaimPolicy to Retain. This ensures that data is retained when a Persistent Volume Claim is removed.

spec.shardPodSpec.podTemplate

Type: collection

Template for the Kubernetes Pods that the MongoDB Enterprise Kubernetes Operator creates for each sharded cluster shard member.

Template values take precedence over values specified in spec.shardPodSpec.

Note

The Kubernetes Operator doesn't validate the fields you provide in spec.shardPodSpec.podTemplate.

spec.shardPodSpec.podTemplate.affinity.podAffinity

Type: string

Kubernetes rule to determine whether multiple MongoDB resource Pods must be co-located with other Pods. To learn more about the use cases, see Affinity and Anti-Affinity in the Kubernetes documentation.

spec.shardPodSpec.podTemplate.affinity.nodeAffinity

Type: string

Kubernetes rule to place Pods for replica set on a specific range of nodes.

For optimized read-write performance, use node affinity rules that restrict Pods to run on particular nodes, or to prefer to run on particular nodes.

spec.shardPodSpec.podTemplate.affinity.podAntiAffinity

Type: string

Default: kubernetes.io/hostname

Sets a rule to spread Pods hosting MongoDB resource to different locations. A location can be a single node, rack, or region. By default, Kubernetes Operator tries to spread pods across different nodes.

spec.shardPodSpec.podTemplate.affinity.podAntiAffinity.requiredDuringSchedulingIgnoredDuringExecution.topologyKey

Type: string

Default: kubernetes.io/hostname

This key defines which label is used to determine which topology domain a node belongs to.

spec.shardPodSpec.podTemplate.metadata

Type: collection

Metadata for the Kubernetes Pods that the MongoDB Enterprise Kubernetes Operator creates for each sharded cluster shard member.

To review which fields you can add to spec.shardPodSpec.podTemplate.metadata, see the Kubernetes documentation.

spec.shardPodSpec.podTemplate.spec

Type: collection

Specifications of the Kubernetes Pods that the MongoDB Enterprise Kubernetes Operator creates for each sharded cluster shard member.

To review which fields you can add to spec.shardPodSpec.podTemplate.spec, see the Kubernetes PodSpec v1 core API.

Note

When you add containers to spec.shardPodSpec.podTemplate.spec.containers, the Kubernetes Operator adds them to the Kubernetes pod. These containers are appended to each sharded cluster shard member containers in the pod.

Use this setting to specify the CPU and RAM allocations for each pod. For examples, see the samples on GitHub.

spec.shardSpecificPodSpec

Type: array

List that contains StatefulSet overrides per shard.

spec.shardSpecificPodSpec.persistence.single

Type: collection

Has Kubernetes Operator create one Persistent Volume Claim and mount all three directories for data, journal, and logs to the same Persistent Volume.

Note

• You must set the values in this collection if spec.persistent : true.

• You may set this collection or the persistence.multiple collections but not both.

Scalar	Data Type	Description
labelSelector	string	Tag used to bind mounted volumes to directories.
storage	string	Minimum size of Persistent Volume that should be mounted. This value is expressed as an integer followed by a unit of storage in JEDEC notation. Default value is 16Gi. Example If the specific shard in requires 60 gigabytes of storage space, set this value to 60Gi.
storageClass	string	Type of storage specified in a Persistent Volume Claim. You may create this storage type as a StorageClass object before using it in this object specification. Note Make sure to set the StorageClass reclaimPolicy to Retain. This ensures that data is retained when a Persistent Volume Claim is removed.

spec.shardSpecificPodSpec.persistence.multiple.data

Type: collection

Has Kubernetes Operator create a Persistent Volume Claim and mount a directory for data to its own Persistent Volume.

Note

• You must set the values in this collection if spec.persistent : true.

• You may set this collection or the persistence.single collection but not both.

Scalar	Data Type	Description
labelSelector	string	Tag used to bind mounted volumes to directories.
storage	string	Minimum storage capacity that must be available on a Kubernetes node to host the specific shard on Kubernetes. This value is expressed as an integer followed by a unit of storage in JEDEC notation. Default value is 16Gi. Example If this MongoDB resource requires 60 gigabytes of storage space, set this value to 60Gi.
storageClass	string	Type of storage needed for the specific shard. You may create this storage type as a StorageClass object before using it in this object specification. Note Make sure to set the StorageClass reclaimPolicy to Retain. This ensures that data is retained when a Persistent Volume Claim is removed.

spec.shardSpecifcPodSpec.persistence.multiple.journal

Type: collection

Has Kubernetes Operator create a Persistent Volume Claim and mount a directory for journal to its own Persistent Volume.

Note

• You must set the values in this collection if spec.persistent : true.

• You may set this collection or the persistence.single collection but not both.

Scalar	Data Type	Description
labelSelector	string	Tag used to bind mounted volumes to directories.
storage	string	Minimum storage capacity that must be available on a Kubernetes node to host the specific shard on Kubernetes. This value is expressed as an integer followed by a unit of storage in JEDEC notation. Default value is 1Gi. Example If this MongoDB resource requires 60 gigabytes of storage space, set this value to 60Gi.
storageClass	string	Type of storage needed for the specific shard. You may create this storage type as a StorageClass object before using it in this object specification. Note Make sure to set the StorageClass reclaimPolicy to Retain. This ensures that data is retained when a Persistent Volume Claim is removed.

spec.shardSpecificPodSpec.persistence.multiple.logs

Type: collection

Has Kubernetes Operator create a Persistent Volume Claim and mount a directory for logs to its own Persistent Volume.

Note

• You must set the values in this collection if spec.persistent : true.

• You may set this collection or the persistence.single collection but not both.

Scalar	Data Type	Description
labelSelector	string	Tag used to bind mounted volumes to directories.
storage	string	Minimum storage capacity that must be available on a Kubernetes node to host the specific shard on Kubernetes. This value is expressed as an integer followed by a unit of storage in JEDEC notation. Default value is 3Gi. Example If this MongoDB resource requires 60 gigabytes of storage space, set this value to 60Gi.
storageClass	string	Type of storage needed for the specific shard. You may create this storage type as a StorageClass object before using it in this object specification. Note Make sure to set the StorageClass reclaimPolicy to Retain. This ensures that data is retained when a Persistent Volume Claim is removed.

spec.shardSpecificPodSpec.podTemplate

Type: collection

Template for the Kubernetes Pods that the MongoDB Enterprise Kubernetes Operator creates for the specific shard.

Template values take precedence over values specified in spec.shardSpecificPodSpec.

Note

The Kubernetes Operator doesn't validate the fields you provide in spec.shardSpecificPodSpec.podTemplate.

spec.shardSpecificPodSpec.podTemplate.affinity.podAffinity

Type: string

Kubernetes rule to determine whether multiple MongoDB resource Pods must be co-located with other Pods. To learn more about the use cases, see Affinity and Anti-Affinity in the Kubernetes documentation.

spec.shardSpecificPodSpec.podTemplate.affinity.podAntiAffinity

Type: string

Default: kubernetes.io/hostname

Sets a rule to spread Pods hosting MongoDB resource to different locations. A location can be a single node, rack, or region. By default, Kubernetes Operator tries to spread pods across different nodes.

spec.shardSpecificPodSpec.podTemplate.affinity.podAntiAffinity.requiredDuringSchedulingIgnoredDuringExecution.topologyKey

Type: string

Default: kubernetes.io/hostname

This key defines which label is used to determine which topology domain a node belongs to.

spec.shardSpecificPodSpec.podTemplate.metadata

Type: collection

Metadata for the Kubernetes Pods that the MongoDB Enterprise Kubernetes Operator creates for the specific shard.

To review which fields you can add to spec.shardSpecificPodSpec.podTemplate.metadata, see the Kubernetes documentation.

spec.shardSpecificPodSpec.podTemplate.spec

Type: collection

Specifications of the Kubernetes Pods that the MongoDB Enterprise Kubernetes Operator creates for the specific shard.

To review which fields you can add to spec.shardSpecificPodSpec.podTemplate.spec, see the Kubernetes PodSpec v1 core API.

Note

When you add containers to spec.shardSpecificPodSpec.podTemplate.spec.containers, the Kubernetes Operator adds them to the Kubernetes pod. These containers are appended to the specific shard containers in the pod.

Use this setting to specify the CPU and RAM allocations for each pod. For examples, see the samples on GitHub.