On this page

Show all

Help Center/ MapReduce Service/ User Guide (Ankara Region)/ Alarm Reference/ ALM-45645 Pending Flush Size of RocksDB Continuously Exceeds the Threshold

ALM-45645 Pending Flush Size of RocksDB Continuously Exceeds the Threshold

Updated on 2024-11-29 GMT+08:00

View PDF

Alarm Description

The system checks the RocksDB monitoring data of jobs at the user-specified alarm reporting interval (metrics.reporter.alarm.job.alarm.rocksdb.metrics.duration, 180s by default). This alarm is generated when the number of pending flush requests of RocksDB for a job continuously reaches n times the number of flush/compaction threads. This alarm is cleared when the number of pending flush requests of RocksDB for the job is less than or equal to the threshold.

The number of flush/compaction threads is the value of state.backend.rocksdb.thread.num. The default value is 2. If SPINNING_DISK_OPTIMIZED_HIGH_MEM is enabled, the default value is 4.
The metrics.reporter.alarm.job.alarm.rocksdb.background.jobs.multiplier parameter specifies n times the number of flush/compaction threads. The default value is 2.

Alarm Attributes

Alarm ID	Alarm Severity	Alarm Type	Service Type	Auto Cleared
45645	Minor	Quality of service	Flink	Yes

Alarm Parameters

Type	Parameter	Description
Location Information	Source	Specifies the cluster for which the alarm was generated.
	ServiceName	Specifies the service for which the alarm was generated.
	ApplicationName	Specifies the name of the application for which the alarm was generated.
	RoleName	Specifies the role for which the alarm was generated.
	JobName	Specifies the job for which the alarm was generated.
Additional Information	ThreshHoldValue	Specifies the threshold value for triggering the alarm.
Additional Information	CurrentValue	Specifies the value that triggered the alarm.

Impact on the System

The checkpoint performance of Flink jobs are affected. There is no impact on the FlinkServer.

Possible Causes

The number of pending flush requests of RocksDB for the Flink job is too large.

Handling Procedure

Check TaskManager logs for the number of pending flush requests and collect logs.

Log in to FusionInsight Manager as a user who has the FlinkServer management permission.
Choose O&M > Alarm > Alarms > ALM-45645 Pending Flush Size of RocksDB Continuously Exceeds the Threshold, view Location, and obtain the name of the task for which the alarm is generated.
Choose Cluster > Services > Yarn and click the link next to ResourceManager WebUI to go to the native Yarn page.

Locate the abnormal task based on its name displayed in Location, search for and record the application ID of the job, and check whether the job logs are available on the Yarn page.

Figure 1 Application ID of a job
- If yes, go to 5.
- If no, go to 6.
Click the application ID of the failed job to go to the job page.
1. Click Logs in the Logs column to view JobManager logs.
  Figure 2 Clicking Logs
2. Click the ID in the Attempt ID column and click Logs in the Logs column to view and save TaskManager logs. Then go to 7.
  Figure 3 Clicking the ID in the Attempt ID column
  
  Figure 4 Clicking Logs
  
  NOTE:
  
  You can also log in to Manager as a user who has the management permission for the current Flink job. Choose Cluster > Services > Flink, and click the link next to Flink WebUI. On the displayed Flink web UI, click Job Management, click More in the Operation column, and select Job Monitoring to view TaskManager logs.

If logs are unavailable on the Yarn page, download logs from HDFS.

On Manager, choose Cluster > Services > HDFS, click the link next to NameNode WebUI to go to the HDFS page, choose Utilities > Browse the file system, and download logs in the /tmp/logs/Username/bucket-logs-tfile/Last four digits of the task application ID/Application ID of the task directory.

Check whether there are too many pending flush requests.

Check whether the sum of the values of rocksdb.mem-table-flush-pending and rocksdb.compaction-pending in TaskManager monitoring logs (keyword RocksDBMetricPrint) is greater than n times the number of RocksDB threads (metrics.reporter.alarm.job.alarm.rocksdb.background.jobs.multiplier, 2 by default). If it is, you can increase the number of RocksDB threads.

If yes, adjust the values of the following custom parameters on the job development page of the Flink web UI, save the settings, and go to 8.

**Table 1** Custom parameters
Parameter	Default Value	Description
state.backend.rocksdb.thread.num	2 4: enables SPINNING_DISK_OPTIMIZED_HIGH_MEM.	Number of flush threads. Increase the number of threads to quickly flush memory data to disks. When the number of threads is increased, the number of vCores also needs to be increased. 2 to 10 are recommended.

If no, go to 9.

Restart the job and check whether the alarm is cleared.
- If yes, no further action is required.
- If no, go to 9.
Contact O&M personnel and send the collected logs.

Alarm Clearance

This alarm is automatically cleared after the fault is rectified.

Related Information

None.

Parent topic: Alarm Reference

Previous topic: ALM-45644 Number of SST Files at Level 0 of RocksDB Continuously Exceeds the Threshold

Next topic: ALM-45646 Pending Compaction Size of RocksDB Continuously Exceeds the Threshold

Feedback

Was this page helpful?

Helpful Not helpful

Provide feedback

Thank you very much for your feedback. We will continue working to improve the documentation.See the reply and handling status in My Cloud VOC.

The system is busy. Please try again later.

Which of the following issues have you encountered?

Content is inconsistent with the product UI

Unclear descriptions

Lack of examples or code

Incorrect steps

Can't find what I need

Lack of best practices

Feedback (optional)

0/500

Select at least one type of issue, and enter your comments or suggestions.

Enter a maximum of 500 characters.

Submit Cancel

For any further questions, feel free to contact us through the chatbot.

Chatbot