Installation Guide

Big Data and Cumulus ® Linux ® Validated Design Guide Deploying Apache Hadoop with Network Switches Running Cumulus ® Linux ®

• Cover
• Contents
• Big Data with Cumulus Linux
• Objective
• Network Demands of Apache Hadoop
• The Power of Open Networking
• Running Apache Hadoop with Cumulus Linux
• Scaling Out
• Converged Administration
• Intended Audience for Network Design and Build
• Building a Data Center Network with Cumulus Linux to Run Apache Hadoop
• Design Steps
• 1. Diagram your Network Architecture.
• 2. Determine which IP Address Each Node Uses
• 3. Determining How Your Cluster Connects to Your Core Network
• 4. Determine the IP Address That Your Servers Use
• 5. Select Master and Backup Systems
• Build Steps
• 1. Set Up Physical Network
• 2. Set Up Basic Configuration of All Switches
• 3. Configure Leaf Switches
• Loopback Configuration
• Spine Port Configuration
• VLAN Configuration
• Enable Quagga and Configure Routing
• 4. Configure Spine Switches
• Loopback Configuration
• Spine Port Configuration
• Enable Quagga and Configure Routing
• 5. Configure the Edge Leaf Switches
• Loopback Configuration
• Spine Port Configuration
• Core Port Configuration
• Enable Quagga and Configure Routing
• 6. Verify Configuration
• 7. Connect Network Fabric
• 8. Install Base OS for Hadoop Nodes
• 9. Ensure Common Services Are Set Up
• 10. Set Up Hadoop
• Conclusion
• Summary
• References
• Appendix: Network Setup Checklist