com.github.mjreid.flinkwrapper
cancelJob cancels an in progress job.
cancelJob cancels an in progress job.
Note that even if the job ID does not exist or is not in a cancellable state, this still returns a success.
cancelJob cancels an in progress job with a savepoint.
cancelJob cancels an in progress job with a savepoint.
If a target directory is supplied, it is used; otherwise, Flink defaults to the directory configured on the server.
This method is asynchronous, on the Flink side; saving the state may take an extended period of time. The CancelJobAccepted.location can be fed into getCancellationStatus() to query the status of the cancellation.
close terminates the actor system and closes the underlying HTTP client.
close terminates the actor system and closes the underlying HTTP client.
This method must be called to properly clean up this client.
getCancellationStatus returns the status of a cancellation that is in progress (i.e.
getCancellationStatus returns the status of a cancellation that is in progress (i.e. as the result of cancelJobWithSavepoint().
getClusterOverview returns the overview of the Flink cluster.
getClusterTaskManagers gets a list of all task managers.
getConfig returns the system level configuration level of the Flink server.
getJobDetails returns detailed information about a single job.
getJobDetails returns detailed information about a single job.
If the job does not exist, None is returned in the future.
getJobExceptions returns all exceptions associated with the job.
getJobExceptions returns all exceptions associated with the job.
If the job does not exist, None is returned in the future.
getJobManagerConfig returns the job manager configurations of the Flink server.
getJobOverview provides a list of all running and finished jobs with a medium level of detail.
getJobPlan returns the job plan JSON for a given job.
getJobPlan returns the job plan JSON for a given job.
If the job does not exist, None is returned in the future.
getJobsList gets a list of all jobs, separated by the state of each job.
runProgram starts a job on the Flink server.
runProgram starts a job on the Flink server.
IMPORTANT - The jarId is *not* the same as what appears in the Flink web UI -- there are hidden GUID values prepended to the JAR name. If you use the uploadJar method, the correct value will be returned in UploadJarResult.filename.
uploadJar uploads a JAR to the Flink server.
FlinkRestClient is the primary contact point for the Flink REST server.
While this class can be invoked manually with a custom client configuration, it's recommended to use one of the factory methods on the FlinkRestClient object to create a client instance.