Apache Flink Submit
CertifiedSubmit Flink job from JAR over REST
Apache Flink Submit
Submit Flink job from JAR over REST
Uploads a JAR to the Flink REST API then runs its entry class. Supports program arguments, explicit parallelism, and savepoint restore. Large JAR uploads are buffered in memory.
type: io.kestra.plugin.flink.SubmitExamples
Submit a Flink batch job
id: flink-batch-job
namespace: company.team
tasks:
- id: submit-job
type: io.kestra.plugin.flink.Submit
restUrl: "http://flink-jobmanager:8081"
jarUri: "s3://flink/jars/my-batch-job.jar"
entryClass: "com.example.BatchJobMain"
args:
- "--input"
- "s3://input/data/"
- "--output"
- "s3://output/results/"
parallelism: 4
Properties
entryClass *Requiredstring
Main class to execute
Fully qualified entry class inside the JAR.
jarUri *Requiredstring
JAR URI
URI of the job JAR to upload. Supports file://, kestra://, s3://, http:// and other schemes. Large artifacts are read fully into memory during upload.
restUrl *Requiredstring
Flink REST API URL
Base URL of the Flink cluster REST API (e.g., http://flink-jobmanager: 8081).
allowNonRestoredState booleanstring
falseAllow non-restored state
Skip savepoint state that cannot be restored; defaults to false and used only when restoreFromSavepoint is set.
args array
Program arguments
Arguments passed to the main method; rendered then space-joined.
jobConfig object
Job configuration
Additional Flink configuration key/values merged into the run request.
parallelism integerstring
Job parallelism
Parallelism to use for this job; defaults to the Flink cluster setting when absent.
pluginDefaultsRef Non-dynamicstring
Reference (ref) of the pluginDefaults to apply to this task.
restoreFromSavepoint string
Restore from savepoint
Path to a savepoint used to start the job statefully.
Outputs
jarId string
Uploaded JAR ID
Identifier of the uploaded JAR on the Flink cluster.
jobId string
Flink job ID
Identifier returned by Flink for the submitted job.
