The HPE Data Fabric Cluster Administration training course provides the knowledge and skills required to plan, install, maintain, and manage a secure Data Fabric cluster, then manage YARN jobs. Learn how to design and install a cluster, and perform pre- and post-installation testing. You configure users and groups, and work with key features of a Data Fabric cluster, including volumes, snapshots, and mirrors-including how to use remote mirrors for disaster recovery. The course also covers monitoring and maintaining disks and nodes, and troubleshooting basic cluster problems. Finally, you install YARN services and practice configuring logging and job schedulers.
By attending HPE Data Fabric Cluster Administration workshop, delegates will learn to:
- Audit and prepare cluster hardware prior to installation
- Run pre-installation tests to verify performance
- Plan a service layout according to cluster configuration and business needs
- Describe the primary architectural components of a HPE Data Fabric installation (nodes, storage pools, volumes, containers, chunks, blocks)
- Use the UI installer to install the HPE Data Fabric distribution
- Define and implement an appropriate node topology
- Define and implement an appropriate volume topology
- Set permissions and quotas for users and groups
- Set up email and alerts
- Set up log aggregation for YARN
- Locate and manipulate configuration files used by the cluster
- Start and stop services
- Use Hadoop commands to perform basic functions
- Use maprcli commands to perform basic functions
- Use the MCS
- Assist with data ingestion
- Configure, monitor, and respond to alerts
- Detect and replace failed disks
- Detect and replace failed nodes
- Create and delete snapshots using both maprcli and the MCS
- Create and delete mirrors using both maprcli and the MCS
- Use mirrors and snapshots for data protection
- Create and implement a disaster recovery plan
- Add, remove, and upgrade ecosystem components
- Monitor and tune job performance
- Configure appropriate job scheduling (FIFO, Fair Scheduler, Capacity Scheduler, Label-base scheduling, Express Lane)
- Set up NFS access to the cluster
- Basic Hadoop knowledge and intermediate Linux knowledge
- Experience using a Linux text editor such as vi
- Familiarity with the Linux command line options such as mv, cp, ssh, grep, and user add
- This HPE Data Fabric Cluster Administration class is ideal for System Administrators who will be creating and maintaining a Hadoop cluster environment.