We will be covering the following topics as part of day to day operations in enterprise clusters.
- Rebalance the cluster -HDFS provides a balancer utility to analyze the block placement since data might not be always placed uniformly across the DataNodes.
- Set up alerting for excessive disk fill – In this we will see how to enable the alerts for any service you would like to receive alerts.
- Define and install a rack topology script – Here we will see how to configure rack topology script.
- Install new type of I/O compression library in cluster – Compression reduces the space needed to store, here we will see how to install new compression techniques other than default that are available.
- Commission/decommission a node – When we need to remove a node from cluster, we should go for decommission first not just deleting. In this topic we will see how to decommission and recommission a node from the cluster.