Shulou(Shulou.com)06/01 Report--

Because the development environment needs to be migrated, the tables from the original HBase are serialized to hdfs using the org.apache.hadoop.hbase.mapreduce.Export tool, downloaded and imported in the new development environment. Two problems were encountered during import.

Problem 1: import failure prompt User hbase cannot submit applications to queue root.hbase

Solution:

Log in to the / etc/hadoop path on the yarn cluster management node Resource Manager machine and modify the fair-scheduler.xml

Vi / etc/hadoop/fair-scheduler.xml, add the following (all Resource Manager nodes need to be modified), save and exit. No need to restart clusters and components

1 drf hdfs,hbase,hive hbase 100 mb,10 vcores 10240 mb,50 vcores 50

Problem 2: the import failed, the terminal did not follow the normal progress, and there was no related error prompt.

Look at the job log in yarn and find the following information

Solution:

Go to the system to check that the id of hbase is uid=482 (hbase), and one of the configuration items of yarn is id, the minimum user of "min.user.id". 1000 is configured here. Fix the problem after changing it to 300.

The purpose of this configuration is to "ID the minimum Linux user allowed. Used to block other superusers."

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.