Flink Job Details¶
After creating a job, you can view the job details to learn about the following information:
Viewing Job Details¶
This section describes how to view job details. After you create and save a job, you can click the job name to view job details, including SQL statements and parameter settings. For a Jar job, you can only view its parameter settings.
In the left navigation pane of the DLI management console, choose Job Management > Flink Jobs. The Flink Jobs page is displayed.
Click the name of the job to be viewed. The Job Detail tab is displayed.
In the Job Details tab, you can view SQL statements, configured parameters.
The following uses a Flink SQL job as an example.
Table 1 Description¶ Parameter
Description
Type
Job type, for example, Flink SQL
Name
Flink job name
Description
Description of a Flink job
Status
Running status of a job
Running Mode
If your job runs on a shared queue, this parameter is Shared.
If your job runs on a custom queue with dedicated resources, this parameter is Exclusive.
Flink Version
Version of Flink selected for the job.
Runtime Configuration
Displayed when a user-defined parameter is added to a job
CUs
Number of CUs configured for a job
Job Manager CUs
Number of job manager CUs configured for a job.
Parallelism
Number of jobs that can be concurrently executed by a Flink job
CU(s) per TM
Number of CUs occupied by each Task Manager configured for a job
Slot(s) per TM
Number of Task Manager slots configured for a job
OBS Bucket
OBS bucket name. After Enable Checkpointing and Save Job Log are enabled, checkpoints and job logs are saved in this bucket.
Save Job Log
Whether the job running logs are saved to OBS
Alarm Generation upon Job Exception
Whether job exceptions are reported
SMN Topic
Name of the SMN topic. This parameter is displayed when Alarm Generation upon Job Exception is enabled.
Auto Restart upon Exception
Whether automatic restart is enabled.
Max. Retry Attempts
Maximum number of retry times upon an exception. Unlimited means the number is not limited.
Savepoint
OBS path of the savepoint
Enable Checkpointing
Whether checkpointing is enabled
Checkpoint Interval
Interval between storing intermediate job running results to OBS. The unit is second.
Checkpoint Mode
Checkpoint mode. Available values are as follows:
At least once: Events are processed at least once.
Exactly once: Events are processed only once.
Idle State Retention Time
Defines for how long the state of a key is retained without being updated before it is removed in GroupBy or Window.
Dirty Data Policy
Policy for processing dirty data. The value is displayed only when there is a dirty data policy. Available values are as follows:
Ignore
Trigger a job exception
Save
Dirty Data Dump Address
OBS path for storing dirty data when Dirty Data Policy is set to Save.
Created
Time when a job is created
Updated
Time when a job was last updated
Checking Job Monitoring Information¶
You can use Cloud Eye to view details about job data input and output.
In the left navigation pane of the DLI management console, choose Job Management > Flink Jobs. The Flink Jobs page is displayed.
Click the name of the job you want. The job details are displayed.
Click Job Monitoring in the upper right corner of the page to switch to the Cloud Eye console.
The following table describes monitoring metrics related to Flink jobs.
Table 2 Monitoring metrics related to Flink jobs¶ Name
Description
Flink Job Data Read Rate
Displays the data input rate of a Flink job for monitoring and debugging. Unit: record/s.
Flink Job Data Write Rate
Displays the data output rate of a Flink job for monitoring and debugging. Unit: record/s.
Flink Job Total Data Read
Displays the total number of data inputs of a Flink job for monitoring and debugging. Unit: records
Flink Job Total Data Write
Displays the total number of output data records of a Flink job for monitoring and debugging. Unit: records
Flink Job Byte Read Rate
Displays the number of input bytes per second of a Flink job. Unit: byte/s
Flink Job Byte Write Rate
Displays the number of output bytes per second of a Flink job. Unit: byte/s
Flink Job Total Read Byte
Displays the total number of input bytes of a Flink job. Unit: byte
Flink Job Total Write Byte
Displays the total number of output bytes of a Flink job. Unit: byte
Flink Job CPU Usage
Displays the CPU usage of Flink jobs. Unit: %
Flink Job Memory Usage
Displays the memory usage of Flink jobs. Unit: %
Flink Job Max Operator Latency
Displays the maximum operator delay of a Flink job. The unit is ms.
Flink Job Maximum Operator Backpressure
Displays the maximum operator backpressure value of a Flink job. A larger value indicates severer backpressure.
0: OK
50: low
100: high
Viewing the Task List of a Job¶
You can view details about each task running on a job, including the task start time, number of received and transmitted bytes, and running duration.
Note
If the value is 0, no data is received from the data source.
In the left navigation pane of the DLI management console, choose Job Management > Flink Jobs. The Flink Jobs page is displayed.
Click the name of the job you want. The job details are displayed.
On Task List and view the node information about the task.
View the operator task list. The following table describes the task parameters.
Table 3 Parameter description¶ Parameter
Description
Name
Name of an operator.
Duration
Running duration of an operator.
Max Concurrent Jobs
Number of parallel tasks in an operator.
Task
Operator tasks are categorized as follows:
The digit in red indicates the number of failed tasks.
The digit in light gray indicates the number of canceled tasks.
The digit in yellow indicates the number of tasks that are being canceled.
The digit in green indicates the number of finished tasks.
The digit in blue indicates the number of running tasks.
The digit in sky blue indicates the number of tasks that are being deployed.
The digit in dark gray indicates the number of tasks in a queue.
Status
Status of an operator task.
Back Pressure Status
Working load status of an operator. Available options are as follows:
OK: indicates that the operator is in normal working load.
LOW: indicates that the operator is in slightly high working load. DLI processes data quickly.
HIGH: indicates that the operator is in high working load. The data input speed at the source end is slow.
Delay
Duration from the time when source data starts being processed to the time when data reaches the current operator. The unit is millisecond.
Sent Records
Number of data records sent by an operator.
Sent Bytes
Number of bytes sent by an operator.
Received Bytes
Number of bytes received by an operator.
Received Records
Number of data records received by an operator.
Started
Time when an operator starts running.
Ended
Time when an operator stops running.
Viewing the Job Execution Plan¶
You can view the execution plan to understand the operator stream information about the running job.
In the left navigation pane of the DLI management console, choose Job Management > Flink Jobs. The Flink Jobs page is displayed.
Click the name of the job you want. The job details are displayed.
Click the Execution Plan tab to view the operator flow direction.
Click a node. The corresponding information is displayed on the right of the page.
Scroll the mouse wheel to zoom in or out.
The stream diagram displays the operator stream information about the running job in real time.
Viewing Job Submission Logs¶
You can view the submission logs to locate the fault.
In the left navigation pane of the DLI management console, choose Job Management > Flink Jobs. The Flink Jobs page is displayed.
Click the name of the job you want. The job details are displayed.
In the Commit Logs tab, view the information about the job submission process.
Viewing Job Running Logs¶
You can view the run logs to locate the faults occurring during job running.
In the left navigation pane of the DLI management console, choose Job Management > Flink Jobs. The Flink Jobs page is displayed.
Click the name of the job you want. The job details are displayed.
On the Run Log tab page, you can view the Job Manager and Task Manager information of the running job.
Information about JobManager and TaskManager is updated every minute. Only run logs of the last minute are displayed by default.
If you select an OBS bucket for saving job logs during the job configuration, you can switch to the OBS bucket and download log files to view more historical logs.
If the job is not running, information on the Task Manager page cannot be viewed.