[FLINK-14376] Introduce the Executor abstraction (FLIP-73)

• Type: New Feature
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 1.10.0
• Fix Version/s: 1.10.0
• Component/s: Client / Job Submission
• Assignee: Kostas Kloudas
• Reporter: Kostas Kloudas
• Created: 11/Oct/19 12:24
• Updated: 09/Dec/19 14:59
• Resolved: 09/Dec/19 14:59

Description

This is the umbrella issue to track the progress of the implementation of FLIP-73.

Issue Links

• relates to: FLINK-14392 Introduce JobClient API(FLIP-74)

Sub-Tasks

• Translate ProgramOptions relevant for job execution to ConfigOptions.: Closed
• Move ClusterDescriptor/ClusterSpecification creation from the CustomCommandLine to a ClusterClientFactory: Closed
• Make ClusterDescriptor/ClusterSpecification creation configuration-based.: Closed
• Refactor ScalaShellRemote(Stream)Environment to simplify inheritance: Closed
• Put job jars from PackagedProgram to Configuration: Closed
• Add Executor-related interfaces and utilities: Closed
• Use new Executor interface in SQL cli: Closed
• Refactor Executor interface and introduce a minimal JobClient interface: Closed
• Make the (Stream)ContextEnvironment use the Executors: Closed
• Asynchronously deploy cluster in AbstractJobClusterExecutor#execute: Closed
• Implement LocalExecutor as new Executor interface: Closed
• Implement RemoteExecutor as a new Executor: Closed
• Remove unused RemoteEnvirnment.executeRemotely() (FLINK-11048): Closed
• Make flink-scala-shell use Executors: Closed
• Remove PlanExecutor and its subclasses: Closed
• Make FlinkYarnCli#isActive() & #getApplicationId() respect config APPLICATION_ID: Closed
• Add public constructors for execution environments that takes Configuration: Closed