REST API

Flink has a monitoring API that can be used to query status and statistics of running jobs, as well as recent completed jobs. This monitoring API is used by Flink’s own dashboard, but is designed to be used also by custom monitoring tools.

The monitoring API is a REST-ful API that accepts HTTP requests and responds with JSON data.

Overview

The monitoring API is backed by a web server that runs as part of the JobManager. By default, this server listens at port 8081, which can be configured in flink-conf.yaml via rest.port. Note that the monitoring API web server and the web dashboard web server are currently the same and thus run together at the same port. They respond to different HTTP URLs, though.

In the case of multiple JobManagers (for high availability), each JobManager will run its own instance of the monitoring API, which offers information about completed and running job while that JobManager was elected the cluster leader.

Developing

The REST API backend is in the flink-runtime project. The core class is org.apache.flink.runtime.webmonitor.WebMonitorEndpoint, which sets up the server and the request routing.

We use Netty and the Netty Router library to handle REST requests and translate URLs. This choice was made because this combination has lightweight dependencies, and the performance of Netty HTTP is very good.

To add new requests, one needs to

  • add a new MessageHeaders class which serves as an interface for the new request,
  • add a new AbstractRestHandler class which handles the request according to the added MessageHeaders class,
  • add the handler to org.apache.flink.runtime.webmonitor.WebMonitorEndpoint#initializeHandlers().

A good example is the org.apache.flink.runtime.rest.handler.job.JobExceptionsHandler that uses the org.apache.flink.runtime.rest.messages.JobExceptionsHeaders.

API

The REST API is versioned, with specific versions being queryable by prefixing the url with the version prefix. Prefixes are always of the form v[version_number]. For example, to access version 1 of /foo/bar one would query /v1/foo/bar.

If no version is specified Flink will default to the oldest version supporting the request.

Querying unsupported/non-existing versions will return a 404 error.

There exist several async operations among these APIs, e.g. trigger savepoint, rescale a job. They would return a triggerid to identify the operation you just POST and then you need to use that triggerid to query for the status of the operation.

For (stop-with-)savepoint operations you can control this triggerId by setting it in the body of the request that triggers the operation. This allow you to safely* retry such operations without triggering multiple savepoints.

The retry is only safe until the async operation store duration has elapsed.

JobManager

[OpenAPI specification](http://nightlies.apache.org/flink/flink-docs-release-1.16/docs/ops/rest_api/

/nightlies.apache.org/flink/flink-docs-release-1.16/generated/rest_v1_dispatcher.yml )

The OpenAPI specification is still experimental.

API reference

v1

/cluster
Verb: DELETEResponse code: 200 OK
Shuts down the cluster
/config
Verb: GETResponse code: 200 OK
Returns the configuration of the WebUI.
/datasets
Verb: GETResponse code: 200 OK
Returns all cluster data sets.
/datasets/delete/:triggerid
Verb: GETResponse code: 200 OK
Returns the status for the delete operation of a cluster data set.
Path parameters
  • triggerid - 32-character hexadecimal string that identifies an asynchronous operation trigger ID. The ID was returned then the operation was triggered.
/datasets/:datasetid
Verb: DELETEResponse code: 202 Accepted
Triggers the deletion of a cluster data set. This async operation would return a ‘triggerid’ for further query identifier.
Path parameters
  • datasetid - 32-character hexadecimal string value that identifies a cluster data set.
/jars
Verb: GETResponse code: 200 OK
Returns a list of all jars previously uploaded via ‘/jars/upload’.
/jars/upload
Verb: POSTResponse code: 200 OK
Uploads a jar to the cluster. The jar must be sent as multi-part data. Make sure that the “Content-Type” header is set to “application/x-java-archive”, as some http libraries do not add the header by default. Using ‘curl’ you can upload a jar via ‘curl -X POST -H “Expect:” -F “jarfile=@path/to/flink-job.jar” http://hostname:port/jars/upload‘.
/jars/:jarid
Verb: DELETEResponse code: 200 OK
Deletes a jar previously uploaded via ‘/jars/upload’.
Path parameters
  • jarid - String value that identifies a jar. When uploading the jar a path is returned, where the filename is the ID. This value is equivalent to the id field in the list of uploaded jars (/jars).
/jars/:jarid/plan
Verb: POSTResponse code: 200 OK
Returns the dataflow plan of a job contained in a jar previously uploaded via ‘/jars/upload’. Program arguments can be passed both via the JSON request (recommended) or query parameters.
Path parameters
  • jarid - String value that identifies a jar. When uploading the jar a path is returned, where the filename is the ID. This value is equivalent to the id field in the list of uploaded jars (/jars).
Query parameters
  • program-args (optional): Deprecated, please use ‘programArg’ instead. String value that specifies the arguments for the program or plan
  • programArg (optional): Comma-separated list of program arguments.
  • entry-class (optional): String value that specifies the fully qualified name of the entry point class. Overrides the class defined in the jar file manifest.
  • parallelism (optional): Positive integer value that specifies the desired parallelism for the job.
/jars/:jarid/run
Verb: POSTResponse code: 200 OK
Submits a job by running a jar previously uploaded via ‘/jars/upload’. Program arguments can be passed both via the JSON request (recommended) or query parameters.
Path parameters
  • jarid - String value that identifies a jar. When uploading the jar a path is returned, where the filename is the ID. This value is equivalent to the id field in the list of uploaded jars (/jars).
Query parameters
  • allowNonRestoredState (optional): Boolean value that specifies whether the job submission should be rejected if the savepoint contains state that cannot be mapped back to the job.
  • savepointPath (optional): String value that specifies the path of the savepoint to restore the job from.
  • program-args (optional): Deprecated, please use ‘programArg’ instead. String value that specifies the arguments for the program or plan
  • programArg (optional): Comma-separated list of program arguments.
  • entry-class (optional): String value that specifies the fully qualified name of the entry point class. Overrides the class defined in the jar file manifest.
  • parallelism (optional): Positive integer value that specifies the desired parallelism for the job.
/jobmanager/config
Verb: GETResponse code: 200 OK
Returns the cluster configuration.
/jobmanager/environment
Verb: GETResponse code: 200 OK
Returns the jobmanager environment.
/jobmanager/logs
Verb: GETResponse code: 200 OK
Returns the list of log files on the JobManager.
/jobmanager/metrics
Verb: GETResponse code: 200 OK
Provides access to job manager metrics.
Query parameters
  • get (optional): Comma-separated list of string values to select specific metrics.
/jobmanager/thread-dump
Verb: GETResponse code: 200 OK
Returns the thread dump of the JobManager.
/jobs
Verb: GETResponse code: 200 OK
Returns an overview over all jobs and their current state.
/jobs
Verb: POSTResponse code: 202 Accepted
Submits a job. This call is primarily intended to be used by the Flink client. This call expects a multipart/form-data request that consists of file uploads for the serialized JobGraph, jars and distributed cache artifacts and an attribute named “request” for the JSON payload.
/jobs/metrics
Verb: GETResponse code: 200 OK
Provides access to aggregated job metrics.
Query parameters
  • get (optional): Comma-separated list of string values to select specific metrics.
  • agg (optional): Comma-separated list of aggregation modes which should be calculated. Available aggregations are: “min, max, sum, avg”.
  • jobs (optional): Comma-separated list of 32-character hexadecimal strings to select specific jobs.
/jobs/overview
Verb: GETResponse code: 200 OK
Returns an overview over all jobs.
/jobs/:jobid
Verb: GETResponse code: 200 OK
Returns details of a job.
Path parameters
  • jobid - 32-character hexadecimal string value that identifies a job.
/jobs/:jobid
Verb: PATCHResponse code: 202 Accepted
Terminates a job.
Path parameters
  • jobid - 32-character hexadecimal string value that identifies a job.
Query parameters
  • mode (optional): String value that specifies the termination mode. The only supported value is: “cancel”.
/jobs/:jobid/accumulators
Verb: GETResponse code: 200 OK
Returns the accumulators for all tasks of a job, aggregated across the respective subtasks.
Path parameters
  • jobid - 32-character hexadecimal string value that identifies a job.
Query parameters
  • includeSerializedValue (optional): Boolean value that specifies whether serialized user task accumulators should be included in the response.
/jobs/:jobid/checkpoints
Verb: GETResponse code: 200 OK
Returns checkpointing statistics for a job.
Path parameters
  • jobid - 32-character hexadecimal string value that identifies a job.
/jobs/:jobid/checkpoints/config
Verb: GETResponse code: 200 OK
Returns the checkpointing configuration.
Path parameters
  • jobid - 32-character hexadecimal string value that identifies a job.
/jobs/:jobid/checkpoints/details/:checkpointid
Verb: GETResponse code: 200 OK
Returns details for a checkpoint.
Path parameters
  • jobid - 32-character hexadecimal string value that identifies a job.
  • checkpointid - Long value that identifies a checkpoint.
/jobs/:jobid/checkpoints/details/:checkpointid/subtasks/:vertexid
Verb: GETResponse code: 200 OK
Returns checkpoint statistics for a task and its subtasks.
Path parameters
  • jobid - 32-character hexadecimal string value that identifies a job.
  • checkpointid - Long value that identifies a checkpoint.
  • vertexid - 32-character hexadecimal string value that identifies a job vertex.
/jobs/:jobid/config
Verb: GETResponse code: 200 OK
Returns the configuration of a job.
Path parameters
  • jobid - 32-character hexadecimal string value that identifies a job.
/jobs/:jobid/exceptions
Verb: GETResponse code: 200 OK
Returns the most recent exceptions that have been handled by Flink for this job. The ‘exceptionHistory.truncated’ flag defines whether exceptions were filtered out through the GET parameter. The backend collects only a specific amount of most recent exceptions per job. This can be configured through web.exception-history-size in the Flink configuration. The following first-level members are deprecated: ‘root-exception’, ‘timestamp’, ‘all-exceptions’, and ‘truncated’. Use the data provided through ‘exceptionHistory’, instead.
Path parameters
  • jobid - 32-character hexadecimal string value that identifies a job.
Query parameters
  • maxExceptions (optional): Comma-separated list of integer values that specifies the upper limit of exceptions to return.
/jobs/:jobid/execution-result
Verb: GETResponse code: 200 OK
Returns the result of a job execution. Gives access to the execution time of the job and to all accumulators created by this job.
Path parameters
  • jobid - 32-character hexadecimal string value that identifies a job.
/jobs/:jobid/jobmanager/config
Verb: GETResponse code: 200 OK
Returns the jobmanager’s configuration of a specific job.
Path parameters
  • jobid - 32-character hexadecimal string value that identifies a job.
/jobs/:jobid/jobmanager/environment
Verb: GETResponse code: 200 OK
Returns the jobmanager’s environment variables of a specific job.
Path parameters
  • jobid - 32-character hexadecimal string value that identifies a job.
/jobs/:jobid/jobmanager/log-url
Verb: GETResponse code: 200 OK
Returns the log url of jobmanager of a specific job.
Path parameters
  • jobid - 32-character hexadecimal string value that identifies a job.
/jobs/:jobid/metrics
Verb: GETResponse code: 200 OK
Provides access to job metrics.
Path parameters
  • jobid - 32-character hexadecimal string value that identifies a job.
Query parameters
  • get (optional): Comma-separated list of string values to select specific metrics.
/jobs/:jobid/plan
Verb: GETResponse code: 200 OK
Returns the dataflow plan of a job.
Path parameters
  • jobid - 32-character hexadecimal string value that identifies a job.
/jobs/:jobid/rescaling
Verb: PATCHResponse code: 200 OK
Triggers the rescaling of a job. This async operation would return a ‘triggerid’ for further query identifier.
Path parameters
  • jobid - 32-character hexadecimal string value that identifies a job.
Query parameters
  • parallelism (mandatory): Positive integer value that specifies the desired parallelism.
/jobs/:jobid/rescaling/:triggerid
Verb: GETResponse code: 200 OK
Returns the status of a rescaling operation.
Path parameters
  • jobid - 32-character hexadecimal string value that identifies a job.
  • triggerid - 32-character hexadecimal string that identifies an asynchronous operation trigger ID. The ID was returned then the operation was triggered.
/jobs/:jobid/savepoints
Verb: POSTResponse code: 202 Accepted
Triggers a savepoint, and optionally cancels the job afterwards. This async operation would return a ‘triggerid’ for further query identifier.
Path parameters
  • jobid - 32-character hexadecimal string value that identifies a job.
/jobs/:jobid/savepoints/:triggerid
Verb: GETResponse code: 200 OK
Returns the status of a savepoint operation.
Path parameters
  • jobid - 32-character hexadecimal string value that identifies a job.
  • triggerid - 32-character hexadecimal string that identifies an asynchronous operation trigger ID. The ID was returned then the operation was triggered.
/jobs/:jobid/status
Verb: GETResponse code: 200 OK
Returns the current status of a job execution.
Path parameters
  • jobid - 32-character hexadecimal string value that identifies a job.
/jobs/:jobid/stop
Verb: POSTResponse code: 202 Accepted
Stops a job with a savepoint. Optionally, it can also emit a MAX_WATERMARK before taking the savepoint to flush out any state waiting for timers to fire. This async operation would return a ‘triggerid’ for further query identifier.
Path parameters
  • jobid - 32-character hexadecimal string value that identifies a job.
/jobs/:jobid/taskmanagers/:taskmanagerid/log-url
Verb: GETResponse code: 200 OK
Returns the log url of jobmanager of a specific job.
Path parameters
  • jobid - 32-character hexadecimal string value that identifies a job.
  • taskmanagerid - 32-character hexadecimal string that identifies a task manager.
/jobs/:jobid/vertices/:vertexid
Verb: GETResponse code: 200 OK
Returns details for a task, with a summary for each of its subtasks.
Path parameters
  • jobid - 32-character hexadecimal string value that identifies a job.
  • vertexid - 32-character hexadecimal string value that identifies a job vertex.
/jobs/:jobid/vertices/:vertexid/accumulators
Verb: GETResponse code: 200 OK
Returns user-defined accumulators of a task, aggregated across all subtasks.
Path parameters
  • jobid - 32-character hexadecimal string value that identifies a job.
  • vertexid - 32-character hexadecimal string value that identifies a job vertex.
/jobs/:jobid/vertices/:vertexid/backpressure
Verb: GETResponse code: 200 OK
Returns back-pressure information for a job, and may initiate back-pressure sampling if necessary.
Path parameters
  • jobid - 32-character hexadecimal string value that identifies a job.
  • vertexid - 32-character hexadecimal string value that identifies a job vertex.
/jobs/:jobid/vertices/:vertexid/flamegraph
Verb: GETResponse code: 200 OK
Returns flame graph information for a vertex, and may initiate flame graph sampling if necessary.
Path parameters
  • jobid - 32-character hexadecimal string value that identifies a job.
  • vertexid - 32-character hexadecimal string value that identifies a job vertex.
Query parameters
  • type (optional): String value that specifies the Flame Graph type. Supported options are: “[FULL, ON_CPU, OFF_CPU]”.
/jobs/:jobid/vertices/:vertexid/metrics
Verb: GETResponse code: 200 OK
Provides access to task metrics.
Path parameters
  • jobid - 32-character hexadecimal string value that identifies a job.
  • vertexid - 32-character hexadecimal string value that identifies a job vertex.
Query parameters
  • get (optional): Comma-separated list of string values to select specific metrics.
/jobs/:jobid/vertices/:vertexid/subtasks/accumulators
Verb: GETResponse code: 200 OK
Returns all user-defined accumulators for all subtasks of a task.
Path parameters
  • jobid - 32-character hexadecimal string value that identifies a job.
  • vertexid - 32-character hexadecimal string value that identifies a job vertex.
/jobs/:jobid/vertices/:vertexid/subtasks/metrics
Verb: GETResponse code: 200 OK
Provides access to aggregated subtask metrics.
Path parameters
  • jobid - 32-character hexadecimal string value that identifies a job.
  • vertexid - 32-character hexadecimal string value that identifies a job vertex.
Query parameters
  • get (optional): Comma-separated list of string values to select specific metrics.
  • agg (optional): Comma-separated list of aggregation modes which should be calculated. Available aggregations are: “min, max, sum, avg”.
  • subtasks (optional): Comma-separated list of integer ranges (e.g. “1,3,5-9”) to select specific subtasks.
/jobs/:jobid/vertices/:vertexid/subtasks/:subtaskindex
Verb: GETResponse code: 200 OK
Returns details of the current or latest execution attempt of a subtask.
Path parameters
  • jobid - 32-character hexadecimal string value that identifies a job.
  • vertexid - 32-character hexadecimal string value that identifies a job vertex.
  • subtaskindex - Positive integer value that identifies a subtask.
/jobs/:jobid/vertices/:vertexid/subtasks/:subtaskindex/attempts/:attempt
Verb: GETResponse code: 200 OK
Returns details of an execution attempt of a subtask. Multiple execution attempts happen in case of failure/recovery.
Path parameters
  • jobid - 32-character hexadecimal string value that identifies a job.
  • vertexid - 32-character hexadecimal string value that identifies a job vertex.
  • subtaskindex - Positive integer value that identifies a subtask.
  • attempt - Positive integer value that identifies an execution attempt.
/jobs/:jobid/vertices/:vertexid/subtasks/:subtaskindex/attempts/:attempt/accumulators
Verb: GETResponse code: 200 OK
Returns the accumulators of an execution attempt of a subtask. Multiple execution attempts happen in case of failure/recovery.
Path parameters
  • jobid - 32-character hexadecimal string value that identifies a job.
  • vertexid - 32-character hexadecimal string value that identifies a job vertex.
  • subtaskindex - Positive integer value that identifies a subtask.
  • attempt - Positive integer value that identifies an execution attempt.
/jobs/:jobid/vertices/:vertexid/subtasks/:subtaskindex/metrics
Verb: GETResponse code: 200 OK
Provides access to subtask metrics.
Path parameters
  • jobid - 32-character hexadecimal string value that identifies a job.
  • vertexid - 32-character hexadecimal string value that identifies a job vertex.
  • subtaskindex - Positive integer value that identifies a subtask.
Query parameters
  • get (optional): Comma-separated list of string values to select specific metrics.
/jobs/:jobid/vertices/:vertexid/subtasktimes
Verb: GETResponse code: 200 OK
Returns time-related information for all subtasks of a task.
Path parameters
  • jobid - 32-character hexadecimal string value that identifies a job.
  • vertexid - 32-character hexadecimal string value that identifies a job vertex.
/jobs/:jobid/vertices/:vertexid/taskmanagers
Verb: GETResponse code: 200 OK
Returns task information aggregated by task manager.
Path parameters
  • jobid - 32-character hexadecimal string value that identifies a job.
  • vertexid - 32-character hexadecimal string value that identifies a job vertex.
/jobs/:jobid/vertices/:vertexid/watermarks
Verb: GETResponse code: 200 OK
Returns the watermarks for all subtasks of a task.
Path parameters
  • jobid - 32-character hexadecimal string value that identifies a job.
  • vertexid - 32-character hexadecimal string value that identifies a job vertex.
/overview
Verb: GETResponse code: 200 OK
Returns an overview over the Flink cluster.
/savepoint-disposal
Verb: POSTResponse code: 200 OK
Triggers the desposal of a savepoint. This async operation would return a ‘triggerid’ for further query identifier.
/savepoint-disposal/:triggerid
Verb: GETResponse code: 200 OK
Returns the status of a savepoint disposal operation.
Path parameters
  • triggerid - 32-character hexadecimal string that identifies an asynchronous operation trigger ID. The ID was returned then the operation was triggered.
/taskmanagers
Verb: GETResponse code: 200 OK
Returns an overview over all task managers.
/taskmanagers/metrics
Verb: GETResponse code: 200 OK
Provides access to aggregated task manager metrics.
Query parameters
  • get (optional): Comma-separated list of string values to select specific metrics.
  • agg (optional): Comma-separated list of aggregation modes which should be calculated. Available aggregations are: “min, max, sum, avg”.
  • taskmanagers (optional): Comma-separated list of 32-character hexadecimal strings to select specific task managers.
/taskmanagers/:taskmanagerid
Verb: GETResponse code: 200 OK
Returns details for a task manager. “metrics.memorySegmentsAvailable” and “metrics.memorySegmentsTotal” are deprecated. Please use “metrics.nettyShuffleMemorySegmentsAvailable” and “metrics.nettyShuffleMemorySegmentsTotal” instead.
Path parameters
  • taskmanagerid - 32-character hexadecimal string that identifies a task manager.
/taskmanagers/:taskmanagerid/logs
Verb: GETResponse code: 200 OK
Returns the list of log files on a TaskManager.
Path parameters
  • taskmanagerid - 32-character hexadecimal string that identifies a task manager.
/taskmanagers/:taskmanagerid/metrics
Verb: GETResponse code: 200 OK
Provides access to task manager metrics.
Path parameters
  • taskmanagerid - 32-character hexadecimal string that identifies a task manager.
Query parameters
  • get (optional): Comma-separated list of string values to select specific metrics.
/taskmanagers/:taskmanagerid/thread-dump
Verb: GETResponse code: 200 OK
Returns the thread dump of the requested TaskManager.
Path parameters
  • taskmanagerid - 32-character hexadecimal string that identifies a task manager.