The Overview page displays the real-time running information about a job. You can analyze and determine whether a job is healthy and meets your expectations based on the job status.

Go to the Overview page

  1. Go to the Administration page in Realtime Compute for Apache Flink.
    1. Log on to the Realtime Compute development platform.
    2. In the top navigation bar, click Administration.
    3. On the Jobs page that appears, click the target job name under the Job Name field.
  2. At the top of the Job Administration page, click Overview.

Task status

The Task Status section displays the number of tasks in each state. A task can be in one of the following states:
  • Created
  • Running
  • Failed
  • Completed
  • Scheduling
  • Canceling
  • Canceled

Job instantaneous values

Below the Task Status section, you can view the instantaneous values of a job.
Parameter Description
Input TPS The number of blocks that are read from the source table per second. For Log Service, you can package multiple data records into one log group for reading. In this case, the value indicates the number of log groups that are read from the source table per second.
Input RPS The number of data records that are read from the source table per second.
Output RPS The number of data records that are written to the result table per second.
Input BPS The number of blocks that are read from the source table per second. Unit: bytes/s.
Consumed CUs The current compute units (CUs) used by the job.
Start Time The start time of the job.
Runtime The duration for which the job has been running.

Vertex topology

The Vertex Topology section displays the execution graph of the underlying computing logic of Realtime Compute for Apache Flink. Each component represents a task. Each data stream flows from one or more data source tables to one or more data result tables. The data flow resembles an arbitrary directed acyclic graph (DAG). To execute a job more efficiently, the underlying logic of Realtime Compute for Apache Flink chains subtasks to form an operator and chains operators to form a task. Each task is executed in a thread. This feature reduces thread switching, message serialization or deserialization, data swapping in the buffer, and data latency. However, this feature increases the overall throughput. An operator indicates the computational logic. A task is a collection of multiple operators.

  • Display mode

    By default, the Vertex Topology section displays the topology. In the upper-right corner of the Vertex Topology section, you can click List Mode to change the display mode.

  • Task status information
    • Task parameters in view mode VertexIn view mode, the information about the task is displayed in the related box. The following table describes the parameters.
      Parameter Description
      Resource Health Score The resource health score is obtained based on the specified check mechanism. The value indicates the job performance. If the resource health score is less than 60, data is stacked up for the task. This results in poor data processing performance.
      Note If the data processing performance is poor, we recommend that you optimize the performance by manual configuration. For more information, see Optimize performance by manual configuration.
      PARALLEL The parallelism for the task.
      TPS The number of blocks that are read from the source table per second.
      DELAY The processing delay of the task.
      IN_Q The percentage of the input queues for the task.
      OUT_Q The percentage of the output queues for the task.
    • Task parameters in list mode
      At the bottom of the Vertex Topology section, you can view the information about the task in list mode. The following table describes the parameters of the task.
      Parameter Description
      Name The name of the task.
      Status The status of the task.
      In Queue The percentage of the input queues for the task.
      Out Queue The percentage of the output queues for the task.
      Delay(ms) The processing delay of the task.
      TPS The amount of data that is read from input nodes per second.
      Bytes Received The amount of data that is received by the task.
      Records Received The number of data records that are received by the task.
      Bytes Sent The amount of data that is sent from the task.
      Records Sent The number of data records that are sent from the task.
      Task The status of each parallelism for the task.
  • Task thread information

    Click the task node. On the SubTasks tab, view the thread list of the task.

Vertex information

Note The following features are supported only in Realtime Compute for Apache Flink V3.0.0 and later.
  • Display Vertex operator information

    In the Vertex Topology section, click the plus sign (+) in the upper-right corner of a Vertex box to display the Vertex operator information.

  • Display Vertex details

    In the upper-right corner of the Vertex Topology section, click Expand All to display Vertex details.

  • Display the Vertex details page
    In the Vertex Topology section, click the Vertex border or the name in the Vertex list to display the Vertex details page on the right. In the Vertex details page, click the SubTasks tab. On the page that appears, click ID to go to the TaskManager page of the related log. For example, below the ID column, click LOG 0. Vertex Detail