5 Launcher Plugin API
The Launcher Plugin API is a set of requests and responses which may be sent from the Launcher to the Plugin (requests) or from the Plugin to the Launcher (responses). This section describes the API calls supported by this version of the RStudio Launcher Plugin SDK. Documentation for the latest version the Launcher Plugin API may be found here.
For each API call, this section will go into detail on how the request corresponds to methods on the IJobSource interface, or on other components of the SDK.
5.1 Common Fields
The following fields are common to all requests and responses:
Request
| Field Name | Description |
|---|---|
| messageType | The type ID of the message. The ID for a given message type can be found in the section for the given message. |
| requestId | The monotonically increasing request ID for this request. This should be saved by the plugin for use in responses. |
Response
| Field Name | Description |
|---|---|
| messageType | The type ID of the message. The ID for a given message type can be found in the section for the given message. |
| requestId | The ID of the request that this response is answering. This is used by the Launcher to determine which response belongs to which request. |
| responseId | The monotonically increasing response ID for this response. The first response ID must be 0. |
The SDK manages all of these fields for each request and response, so the Plugin developer does not need to worry about them.
In addition, many requests may also include username and requestUsername fields, described in the table below. These fields are passed to the appropriate IJobSource method as they impact the behaviour of the Plugin.
| Field Name | Description |
|---|---|
| username | The username of the user that initiated the request. Used for auditing/security purposes. Some requests have the ability to specify ’*’ for this value, indicating all users’ information should be returned. This can occur if a super user (admin user) makes a request, or if authorization is disabled. |
| requestUsername | The actual username used when the request was submitted. This should be ignored in most plugins, and the username value (above) should be used. This field is provided for additional information if needed. |
5.2 Requests and Responses
5.2.1 Error
For any request, the Plugin may return an Error response. To return an Error response to the server, the IJobSource implementation should return an Error. The Plugin API component will generally construct the appropriate Error response based on the Error returned by the IJobSource call. For the most part, the Plugin developer does not need to be concerned with Error responses.
Error Response
| Field Name | Description | Value |
|---|---|---|
| messageType | See above. | Response::Type::ERROR (-1) |
| requestId | See above. | Int |
| responseId | See above. | Int |
| errorCode | The error code. More details in the Error Codes section. | Int |
| errorMessage | The error message. More details in the Error Codes section. | Int |
5.2.1.1 Error Codes
The code of the Error response must be one of the error codes supported by the Launcher Plugin API. The message can be an arbitrary string which provides more context for the Launcher user.
| Code Name | Description | Value |
|---|---|---|
| UnknownError | The request failed for an undetermined reason. Used when the Plugin cannot determine an appropriate error code for the error. | 0 |
| RequestNotSupported | The request is not supported by the Plugin. The Plugin API component may return this if the Launcher sends a request that is not part of the Plugin’s supported Launcher Plugin API version. | 1 |
| InvalidRequest | The request is malformed. The Plugin API component may return this if it receives an unexpected message from the Launcher. | 2 |
| JobNotFound | The job does not exist in the scheduling system. The Plugin should return this if the user specified job ID does not exist. | 3 |
| PluignRestarted | The request could not be completed because the Plugin had to restart. | 4 |
| Timeout | The request timed out while waiting for a response from the job scheduling system. | 5 |
| JobNotRunning | The job exists in the job scheduling system but is not in the running state. | 6 |
| JobOutputNotFound | The job does not have output. | 7 |
| InvalidJobState | The job has an invalid job state for the requested action. | 8 |
| JobControlFailure | The job control action failed. | 9 |
| UnsupportedVersion | The Launcher is using a Launcher Plugin API version that is not supported by the Plugin. | 10 |
5.2.2 Heartbeat
When the Launcher starts the Plugin, it provides the configured heartbeat-interval-seconds value to the Plugin. If the Plugin does not send the Heartbeat Response every configured amount of time, the Launcher will restart the Plugin. This request is completely handled by the SDK and doesn’t require any action on the part of the Plugin developer. However, when debugging issues in the Plugin, it can become problematic if the Launcher restarts the Plugin while getting the debugger set up. To avoid this, heartbeat-interval-seconds can be set to 0 in /etc/rstudio/launcher.conf, which will disable heartbeating. This is not a recommended configuration for production because unresponsive Plugins will not be handled correctly.
Despite the fact that there will be many heartbeats sent to and from the Launcher, every Heartbeat Request and Response will have request ID and response IDs of 0. This is because it is Heartbeat Response do not strictly respond to a particular request – they must merely be sent to the Launcher every configured time interval.
Heartbeat Request
| Field Name | Description | Value |
|---|---|---|
| messageType | See above. | Request::Type::HEARTBEAT (0) |
| requestId | See above. | 0 |
Heartbeat Response
| Field Name | Description | Value |
|---|---|---|
| messageType | See above. | Response::Type::HEARTBEAT (0) |
| requestId | See above. | 0 |
| responseId | See above. | 0 |
5.2.3 Bootstrap
This request is sent to the Plugin once, immediately after the Plugin is launched. During this request, the SDK will verify version compatibility with the Launcher and then trigger the population of existing jobs by invoking IJobSource::getJobs.
Bootstrap Request
| Field Name | Description | Value |
|---|---|---|
| messageType | See above. | Request::Type::BOOTSTRAP (1) |
| requestId | See above. | 0 |
| version | The Launcher Plugin API version in use by the Launcher. | Version |
Bootstrap Response
| Field Name | Description | Value |
|---|---|---|
| messageType | See above. | Response::Type::BOOTSTRAP (1) |
| requestId | See above. | 0 |
| responseId | See above. | 0 |
| version | The Launcher Plugin API version in use by the Plugin. | Version |
5.2.4 Jobs
There are two API requests which return a Job State Response: the Submit Job and Job State requests. The SDK maintains an internal cache of all jobs in the system. When a Job State request is received, the SDK searches the cache for any jobs that match the request and sends them back to the Launcher. The SDK will also manage filtering the jobs and restricting the job fields in the response, if necessary. For more information about how the job cache is maintained, see the Customizing the Job Repository section of the Advanced Features chapter.
Submit Job Request
| Field Name | Description | Value |
|---|---|---|
| messageType | See above. | Request::Type::SUBMIT_JOB (2) |
| requestId | See above. | Int |
| username | See above. | String |
| requestUsername | See above. | String |
| job | The job object which describes the job to be launched. | Job |
Job State Request
| Field Name | Description | Value |
|---|---|---|
| messageType | See above. | Request::Type::GET_JOB (3) |
| requestId | See above. | Int |
| username | See above. | String |
| requestUsername | See above. | String |
| jobId | The ID of the job to be retrieved. If this field is ’*’ all jobs should be retrieved and filters may be used. | String |
| encodedJobId | The Launcher generated encoded job ID which contains extra metadata. Provides extra information only. | String |
| tags | If present, the set of tags by which to filter the jobs. Only jobs which have all these tags will be returned. | String Array |
| startTime | If present, only jobs which were submitted after this UTC time will be returned. Format: YYYY-MM-DDThh:mm:ss. |
String |
| endTime | If present, only jobs which were submitted before this UTC time will be returned. Format: YYYY-MM-DDThh:mm:ss. |
String |
| statuses | If present, only jobs which have one of the specified statuses will be returned. | JobState Array |
| fields | If present, only the job fields included in this list will be included in the response, excepted ‘id’, which will always be returned. | String Array |
Job State Response
| Field Name | Description | Value |
|---|---|---|
| messageType | See above. | Response::Type::JOB_STATE (2) |
| jobs | The list of jobs that met the request criteria. | Job Array |
5.2.5 Job Status Stream
This API call is responsible for streaming Job status updates as they occur. The SDK maintains an internal cache of all the Jobs in the Job Scheduling System and uses the JobStatusNotifier to allow components of the SDK to be notified when a Job’s status is updated. Components which wish to be notified of a Job status update may subscribe to the notifier, and components which wish to notify other components about new updates should post the updates to the JobStatusNotifier.
The SDK provides AbstractJobStatusWatcher and AbstractTimedJobStatusWatcher to help facilitate the implementation of keeping Job statuses updated within the Plugin. For more details about implementing Job status updates, see the Job Status Updates of the Advanced Features chapter, or TODOs #’s 9 - 11 of the RStudio Launcher Plugin SDK QuickStart Guide.
The SDK will manage all Job status streams, including closing them on request. The Plugin developer is only responsible for posting correct Job status updates to the JobStatusNotifier either directly or indirectly.
Rather than sending a response per open stream, the Plugin sends the Launcher a response with a field that indicates which stream requests are answered by this response. A sequence ID is kept per request to ensure the Launcher can keep the responses in the right order.
Job Status Stream Request
| Field Name | Description | Value |
|---|---|---|
| messageType | See above. | Request::Type::GET_JOB_STATUS (4) |
| requestId | See above. | Int |
| username | See above. | String |
| requestUsername | See above. | String |
| jobId | The ID of the job to be for which to stream statuses. If this field is ’*’ the statuses of all jobs will be streamed. | String |
| encodedJobId | The Launcher generated encoded job ID which contains extra metadata. Provides extra information only. | String |
| cancel | Whether the stream should be started (false) or canceled (true). Default: false. |
Boolean |
Job Status Stream Response
| Field Name | Description | Value |
|---|---|---|
| messageType | See above. | Response::Type::JOB_STATUS (3) |
| sequences | An array of requests which are waiting for this stream response and the sequence ID of this response, relative to each request ID. | StreamSequence Array |
| id | The ID of the job. | String |
| name | The name of the job. | String |
| status | The status of the job. | JobState |
| statusMessage | The status message of the job, if any. | String |
5.2.6 Control Job
This API call can be used to control the state of a job. Currently supported operations are cancel, stop, kill, suspend, and resume. The Job must be in the correct state for a given operation to be valid (respectively, pending, running, running, running, and suspended). The SDK will handle Job State validation. The Plugin is only responsible for a best-effort attempt to perform the given operation on the job. If the Plugin indicates that the operation is not supported, the SDK will return an Error Response.
Control Job Request
| Field Name | Description | Value |
|---|---|---|
| messageType | See above. | Request::Type::CONTROL_JOB (5) |
| requestId | See above. | Int |
| username | See above. | String |
| requestUsername | See above. | String |
| jobId | The ID of the job to control. This field cannot be ’*’. | String |
| encodedJobId | The Launcher generated encoded job ID which contains extra metadata. Provides extra information only. | String |
| operation | The operation to be performed on the Job. | JobOperation |
Control Job Response
| Field Name | Description | Value |
|---|---|---|
| messageType | See above. | Response::Type::CONTROL_JOB (4) |
| requestId | See above. | Int |
| responseId | See above. | Int |
| statusMessage | A message describing the status of the control job operation, if any. | String |
| operationComplete | Whether the control job operation completed successfully or not. | Boolean |
5.2.7 Job Output Stream
This API call is responsible for streaming a Job’s output data. The caller may request the type of output they wish to see (Standard Out, Standard Error, or both) and the Plugin must do its best to report only that type of data.
The SDK will manage all the output streams. The Plugin developer needs to implement IJobSource::createOutputStream and may either use the provided FileOutputStream class or implement a custom concrete AbstractOutputStream. An example using the FileOutputStream can be found in ‘TODO #8’ of the ‘RStudio Launcher Plugin SDK QuickStart Guide’. For more details on implementing a custom AbstractOutputStream class, see the Custom Output Streams section.
Job Output Stream Request
| Field Name | Description | Value |
|---|---|---|
| messageType | See above. | Request::Type::GET_JOB_OUTPUT (6) |
| requestId | See above. | Int |
| username | See above. | String |
| requestUsername | See above. | String |
| jobId | The ID of the job to be for which to stream output. | String |
| encodedJobId | The Launcher generated encoded job ID which contains extra metadata. Provides extra information only. | String |
| outputType | The type of output to be streamed. Standard out (0), Standard error (1), or both (2). | Int: 0, 1, or 2 |
| cancel | Whether the stream should be started (false) or canceled (true). Default: false. |
Boolean |
Job Output Stream Response
| Field Name | Description | Value |
|---|---|---|
| messageType | See above. | Response::Type::JOB_OUTPUT (5) |
| requestId | See above. | Int |
| responseId | See above. | Int |
| seqId | The ID of this response in the stream of output responses. | Int |
| outputType | The type of output to in the response, if any. Standard out, Standard error, or mixed. | String: “stdout”, “stderr”, or “mixed” |
| output | The job output, if any. | String |
| complete | Whether the output stream is complete. | Boolean |
5.2.8 Job Resource Utilization Stream
This API call is responsible for streaming a Job’s resource utilization metrics. Resource utilization metrics may only be streamed when the Job is in the Running state. If the Job is not running, a JobNotRunning error will be returned by the SDK. Otherwise, the SDK will invoke AbstactResourceStream::initialize and the Plugin should commence streaming resource utilization metrics.
The SDK will manage all the resource utilization streams. The Plugin developer needs to implement AbstactResourceStream::initialize and may either extend AbstractResourceStream directly, or extend AbstractTimedResourceStream if resource utilization data must be polled. An example of extending AbstractResourceStream directly can be found in ‘TODO #16’ of the ‘RStudio Launcher Plugin SDK QuickStart Guide’. The LocalResourceStream provided with the sample Local Plugin provides an example of extending AbstractTimedResourceStream.
Job Resource Utilization Stream Request
| Field Name | Description | Value |
|---|---|---|
| messageType | See above. | Request::Type::GET_JOB_RESOURCE_UTIL (7) |
| requestId | See above. | Int |
| username | See above. | String |
| requestUsername | See above. | String |
| jobId | The ID of the job for which to stream resource utilization metrics. | String |
| encodedJobId | The Launcher generated encoded job ID which contains extra metadata. Provides extra information only. | String |
| cancel | Whether the stream should be started (false) or canceled (true). Default: false. |
Boolean |
Job Resource Utilization Stream Response
| Field Name | Description | Value |
|---|---|---|
| messageType | See above. | Response::Type::JOB_RESOURCE_UTIL (6) |
| requestId | See above. | Int |
| responseId | See above. | Int |
| seqId | The ID of this response in the stream of resource utilization responses. | Int |
| cpuPercent | The percentage of CPU utilization of the Job, if available. Optional. | String |
| cpuSeconds | The total run time of the Job, in seconds, if available. Optional. | String |
| virtualMemory | The current size of virtual memory in use by the Job, in MB, if available. Optional. | String |
| residentMemory | The current size of the resident set in use by the Job, in MB, if available. Optional. | String |
| complete | Whether the resource utilization stream is complete. | Boolean |
5.2.9 Job Network
This API call is responsible for providing network information about the machine running the specified job to the caller. When this API call is received, the SDK will invoke IJobSource::getNetworkInfo. The Plugin is responsible for resolving the hostname and the IP addresses of the machine running the job.
For an example of how to implement IJobSource::getNetworkInfo, see ‘TODO #14’ of the ‘RStudio Launcher Plugin SDK QuickStart Guide’.
Job Network Request
| Field Name | Description | Value |
|---|---|---|
| messageType | See above. | Request::Type::GET_JOB_NETWORK (8) |
| requestId | See above. | Int |
| username | See above. | String |
| requestUsername | See above. | String |
| jobId | The ID of the job to be for which to retreive network information. | String |
| encodedJobId | The Launcher generated encoded job ID which contains extra metadata. Provides extra information only. | String |
Job Network Response
| Field Name | Description | Value |
|---|---|---|
| messageType | See above. | Response::Type::GET_JOB_NETWORK (7) |
| requestId | See above. | Int |
| responseId | See above. | Int |
| host | The hostname of the machine running the job. | String |
| ipAddresses | An array of the non-local IP addresses for the machine running the job. | String Array |
5.2.10 Cluster Info
This API call is responsible for reporting the configuration and capabilities of the job scheduling system. RStudio applications use this call to determine what UI to surface to the end user when they launch jobs. When this request is received, the SDK will invoke the following method to compose the correct response:
IJobSource::getConfiguration
The Plugin may return an error from this method if it encounters any issues populating the fields on the JobSourceConfiguration object. Additionally, the method is provided with a system::User object, which represents the user who initiated the Cluster Info request. This optionally allows the Plugin Developer to provide different results on a per-user or per-group basis. This may be useful to allow system administrator to set different maximum and default Resource Limits for different groups of user via the User Profiles feature. It also may be useful if the job scheduling system allows administrators to make similar rule sets.
For a simple example of implementing this method, see ‘TODO #7’ in the ‘RStudio Launcher Plugin SDK QuickStart Guide’.
Cluster Info Request
| Field Name | Description | Value |
|---|---|---|
| messageType | See above. | Request::Type::GET_CLUSTER_INFO (9) |
| requestId | See above. | Int |
| username | See above. | String |
| requestUsername | See above. | String |
Cluster Info Response
| Field Name | Description | Value |
|---|---|---|
| messageType | See above. | Response::Type::CLUSTER_INFO (8) |
| requestId | See above. | Int |
| responseId | See above. | Int |
| supportsContainers | Whether or not the job scheduling system supports containers. | Boolean |
| config | Any custom configuration values which may be set per Job. | JobConfig Array |
| placementConstraints | Any custom placement constraints which may be requested for a Job. | PlacementConstraint Array |
| queues | The queues which are available to run Jobs, if the job scheduling system supports queues. | String Array |
| resourceLimits | The types of resource limits which may be requested for a Job, including default and maximum values, if any. | ResourceLimit Array |
| images | If the job scheduling system supports containers, the list of container images which may be used when launching a job. | String Array |
| defaultImage | If the job scheduling system supports containers, the default container image to use if none is selected. | String |
| allowUnknownImages | If the job scheduling system supports containers, whether to allow Jobs to be run on images which are not known. | Boolean |
5.3 Complex API Types
This section describes the details of API types which are not simple types, such as String, Int, or Boolean.
5.3.1 Container
| Field Name | Description | Value |
|---|---|---|
| image | The name of the container image to use. | String |
| runAsUserId | The ID of the user to run the container as. Optional. | Int |
| runAsGroupId | The ID of the group to run the container as. Optional. | Int |
| supplememtalGroupIds | The list of additional group IDs to be added to the run-as user in the container. Optional. | Int Array |
5.3.2 Environment
| Field Name | Description | Value |
|---|---|---|
| name | The name of the environment variable. | String |
| value | The value of the environment variable. | String |
5.3.3 ExposedPort
| Field Name | Description | Value |
|---|---|---|
| targetPort | The target port, within the container. | Int |
| publishedPort | The published port, if different from the container port. | Int |
| protocol | The network protocol to use. Default: TCP. | String |
5.3.4 Job
| Field Name | Description | Value |
|---|---|---|
| args | The arguments of the ‘command’ or ‘exe’ of the Job. | String Array |
| cluster | The cluster of the Job. | String |
| command | The shell command of the Job. Mutually exclusive with ‘exe’. | String |
| config | The custom configuration values of the Job. | JobConfig Array |
| container | The container configuration of the Job, if the Cluster supports containers. | Container Array |
| environment | The environment variables for the Job. | Environment Array |
| exe | The executable of the Job. Mutually exclusive with ‘command’. | String |
| exitCode | The exit code of the ‘command’ or ‘exe’ of the Job. | Int |
| exposedPorts | The exposed ports of the Job, if containers are used. | ExposedPort Array |
| host | The host on which the Job was (or is being) run. | String |
| id | The unique ID of the Job. | String |
| lastUpdateTime | The time of the last update to the Job. | String (ISO 8601 format DateTime) |
| mounts | The file system mounts to apply when the Job is run. | Mount Array |
| name | The name of the Job. | String |
| pid | The process ID of the Job, if applicable. | Int |
| id | The unique ID of the Job. | String |
| placementConstraints | The list of placement constraints that were selected for the Job. | PlacementConstraint Array |
| queues | The list of queues that may be used to launch the Job, or the queue that was used to run the Job. | String Array |
| stdin | The standard input to be passed to the ‘command’ or ‘exe’ of the Job. | String |
| stderr | The location of the file which contains the standard error output of the Job. | String |
| stdout | The location of the file which contains the standard output of the Job. | String |
| status | The current status of the Job. | JobState |
| statusMessage | The message or reason of the current status of the Job. | String |
| submissionTime | The time at which the Job was submitted to the Cluster. | String (ISO 8601 format DateTime) |
| tags | The tags that were set for the Job. Used for filtering Jobs. | String Array |
| user | The username of the user who launched the Job. | String |
| workingDirectory | The directory to use as the working directory for the ‘command’ or ‘exe’. | String |
5.3.5 JobConfig
| Field Name | Description | Value |
|---|---|---|
| name | The name of the custom configuration value. | String |
| valueType | The type of the custom configuration value. Optional. | “string”, “int”, “float”, or “enum” |
| value | The value of the custom configuration value. Optional. | String (convertible to valueType) |
5.3.6 JobOperation
This is a JSON Integer value with a limited set of values. The possible values of a JobOperation field and their meanings are listed in the table below.
| Value | Description |
|---|---|
| 0 | The Job should be suspended. Only valid if the Job is currently in the “Running” JobState. |
| 1 | The Job should be resumed. Only valid if the Job is currently in the “Suspended” JobState. |
| 2 | The Job should be stopped. Only valid if the Job is currently in the “Running” JobState. |
| 3 | The Job should be killed. Only valid if the Job is currently in the “Running” JobState. |
| 4 | The Job should be canceled. Only valid if the Job is currently in the “Pending” JobState. |
5.3.7 JobState
This type is a JSON String value with limited set of valid values. The possible values of a JobState field and their meanings are listed in the table below.
| Value | Description |
|---|---|
| Canceled | The job was canceled by the user before it began to run. |
| Failed | The job could not be launched due to an error. This status does not refer to jobs where the process exited with a non-zero exit code. |
| Finished | The job was launched and finished executing. This includes jobs where the process exited with a non-zero exit code. |
| Killed | The job was forcibly killed while it was running, i.e. the job process received SIGKILL. |
| Pending | The job was successfully submitted to the job scheduling system but has not started running yet. |
| Running | The job is currently running. |
| Suspended | The job was running, but execution was paused and may be resumed at a later time. |
5.3.8 Mount
| Field Name | Description | Value |
|---|---|---|
| mountPath | The destination path of the mount. | String |
| readOnly | Whether the source path should be mounted with write permissions (false) or not (true). Default: false. |
Boolean |
| type | The type of mount. The default supported options are ‘azureFile’, ‘cephFs’, ‘glusterFs’, ‘host’, and ‘nfs’. The ‘passthrough’ value or a custom value may be used for other mount types. | String |
| source | The mount source description. Must match the specified mount type. | MountSource |
5.3.9 MountSource
This object will have a different schema depending on the type of mount specified in the ‘type’ field of the Mount object. Below, the source object for each of the default supported mount types will be described. If the mount type is ‘passthrough’ or a custom value, the object definition is determined by the Plugin.
5.3.9.1 AzureFileMountSource
This represents an Azure File source. If the ‘type’ field in the Mount object is ‘azureFile’ and the ‘source’ object follows this schema, the Plugin should attempt to mount the requested share from the Azure File account specified in the secret.
| Field Name | Description | Value |
|---|---|---|
| secretName | The name of the secret that contains both the Azure storage account name and the key. | String |
| shareName | The share name to be used. | String |
5.3.9.2 CephFsMountSource
This represents a Ceph File System source. If the ‘type’ field in the Mount object is ‘cephFs’ and the ‘source’ object follows this schema, the Plugin should attempt to mount the requested path as the specified user from the list of Ceph monitor addresses.
| Field Name | Description | Value |
|---|---|---|
| monitors | A comma-separated list of Ceph monitor addresses. For example: 192.168.1.200:8765,192.168.1.200:8766 |
String Array |
| path | The path within the Ceph filesystem to mount. | String |
| user | The Ceph username to use. | String |
| secretFile | The file which contains the Ceph keyring for authentication. | String |
| secretRef | Reference to Ceph authentication secrets, which overrides SecretFile if specified. | String |
5.3.9.3 GlusterFsMountSource
This represents a Glusterfs mount source. If the ‘type’ field in the Mount object is ‘glusterFs’ and the ‘source’ object follows this schema, the Plugin should attempt to mount the requested path from the specified Glusterfs endpoints.
| Field Name | Description | Value |
|---|---|---|
| endpoints | The name of the endpoints object that represents a Gluster cluster configuration. | String |
| path | The name of the GlusterFs volume. | String |
5.3.9.4 HostMountSource
This represents a host mount source. If the ‘type’ field in the Mount object is ‘host’ and the ‘source’ object follows this schema, the Plugin should attempt to mount the requested path from the Launcher host, or reject the request on the basis that mounting host paths is not supported.
| Field Name | Description | Value |
|---|---|---|
| path | The source path to be mounted. | String |
5.3.9.5 NfsMountSource
This represents an NFS server mount source. If the ‘type’ field in the Mount object is ‘nfs’ and the ‘source’ object follows this schema, the Plugin should attempt to mount the requested path from the specified NFS server, or reject the request on the basis that mounting NFS paths is not supported.
| Field Name | Description | Value |
|---|---|---|
| host | The host of the NFS server. | String |
| path | The source path of the mount, on the NFS server. | String |
5.3.10 PlacementConstraint
| Field Name | Description | Value |
|---|---|---|
| name | The name of the placement constraint. | String |
| value | One of the possible values of the placement constraint. Optional. | String |
5.3.11 ResourceLimit
| Field Name | Description | Value |
|---|---|---|
| limitType | The type of the resource. | “cpuCount”, “cpuTime”, “memory”, or “memorySwap” |
| defaultValue | The default value of the resource type. | String |
| maxValue | The maximum value of the resource type. | String |
| value | The requested value of the resource. | String |