HiveServer and HiveHCat Process Faults

Issue

The HiveServer and WebHCat processes in the customer cluster are faulty.

Symptom

The HiveServer and WebHCat processes on the Master2 node in the MRS cluster are faulty. After the restart, the processes are still faulty.

Cause Analysis

On Manager, start the faulty HiveServer process. Log in to the node and search for the error information at the corresponding time point in the hiveserver.out log file. Error messages error parsing conf mapred-site.xml and Premature end of file are found. Restart WebHCat. The same error is reported because the mapred-site.xml file fails to be parsed.

Procedure

1. Log in to the Master2 node as user root.
2. Run the find / -name 'mapred-site.xml' command to obtain the location of the mapred-site.xml file.
   • The path of HiveServer is /opt/Bigdata/Cluster version/1_13_HiveServer/etc/mapred-site.xml.
   • The path of WebHCat is /opt/Bigdata/Cluster version/1_13_WebHCat/etc/mapred-site.xml.
3. Check whether the mapred-site.xml file is normal. In this case, the configuration file is empty. As a result, the parsing fails.
4. Restore the mapred-site.xml file. Run the scp command to copy the configuration file in the corresponding directory on the Master1 node to the corresponding directory on the Master2 node to replace the original file.
5. Run the chown omm:wheel mapred-site.xml command to change the owner group and user.
6. On Manager, restart the faulty HiveServer and WebHCat processes.