5.3.6. Percent DataNodes storage full alert

This alert is triggered if the percentage of DataNodes in storage full condition exceeds the configured critical threshold. It uses the check_aggregate plugin, to aggregate the results of DATANODE::Storage full alert checks.

 5.3.6.1. Potential causes
  • Check for dead DataNodes in the Services list.

  • Data distribution among the live nodes is not balanced

 5.3.6.2. Possible remedies
  • Run the Hadoop Balancer tool to distribute data to relatively less used DataNodes, freeing up space on the full nodes.


loading table of contents...