ALM-19025 Damaged StoreFile in HBase

Alarm Description

The system checks the hdfs://hacluster/hbase/autocorrupt and hdfs://hacluster/hbase/MasterData/autocorrupt directories on HDFS of each HBase service every 120 seconds. This alarm is generated when there are files in the directories.

This alarm is cleared when the hdfs://hacluster/hbase/autocorrupt and hdfs://hacluster/hbase/MasterData/autocorrupt directories do not exist or are empty.

hdfs://hacluster indicates the name of the file system used by HBase, and /hbase indicates the root directory of HBase in the file system. You can log in to FusionInsight Manager, choose Cluster > Services > HBase and click Configuration. Search for fs.defaultFS and hbase.data.rootdir.

Alarm Attributes

Alarm ID	Alarm Severity	Alarm Type	Service Type	Auto Cleared
19025	Major	Error handling	HBase	Yes

Alarm Parameters

Type	Parameter	Description
Location Information	Source	Specifies the cluster for which the alarm is generated.
	ServiceName	Specifies the service for which the alarm is generated.
	RoleName	Specifies the role for which the alarm is generated.
	HostName	Specifies the host for which the alarm is generated.

Impact on the System

Data in the file may be lost, and data queried by the service may be inconsistent.

Possible Causes

The StoreFile files are damaged.

Handling Procedure

Log in to FusionInsight Manager and choose O&M. In the navigation pane on the left, choose Alarm > Alarms. On the page that is displayed, locate the row containing the alarm whose Alarm ID is 19025, and view the service in Location.
Log in to the node where the HDFS and HBase clients are installed as the client installation user and run the following commands:

cd Client installation directory

source bigdata_env

kinit Component service user (If Kerberos authentication is disabled for the cluster (the cluster is in normal mode), skip this step.)
Check the damaged StoreFile file.
- Run the following command to check whether the /hbase/autocorrupt directory of HDFS is empty. If it is not, go to 4.
  hdfs dfs -ls -R hdfs://hacluster/hbase/autocorrupt
- Run the following command to check whether the /hbase/MasterData/autocorrupt directory of HDFS is empty. If it is not, go to 9.
  hdfs dfs -ls -R hdfs://hacluster/hbase/MasterData/autocorrupt
Run the following command to restore the StoreFile files in the hdfs://hacluster/hbase/autocorrupt directory:

hdfs debug recoverLease -path hdfs://hacluster/hbase/autocorrupt/Name space/Table/Region/Column family/StoreFile files
Check whether the damaged StoreFile files are restored. If the following information is displayed, the restoration is successful:
```
recoverLease SUCCEEDED on hdfs://hacluster/hbase/autocorrupt/default/h1/865665fe32db62dadada68b644359809/cf1/95f210f931ad44c99e4028470be7d292
```
If yes, go to 6.

If no, go to 9.
Run the following command to move the files back to the hdfs://hacluster/hbase/data directory:

hdfs dfs -mv hdfs://hacluster/hbase/autocorrupt/Name space/Table/Region/Column family/StoreFile fileshdfs://hacluster/hbase/data/Name space/Table/Region/Column family/StoreFile files
Run the following command on HBase Shell to bring the region online again:

hbase shell

unassign'Region'

assign'Region'
Wait several minutes and check whether the alarm is cleared.
- If yes, no further action is required.
- If no, go to 9.

Collect fault information.

On FusionInsight Manager, choose O&M. In the navigation pane on the left, choose Log > Download.
Expand the Service drop-down list, and select HBase for the target cluster.
Click the edit icon in the upper right corner, and set Start Date and End Date for log collection to 10 minutes ahead of and after the alarm generation time, respectively. Then, click Download.
Contact O&M engineers and provide the collected logs.