Book Image

Linux for Networking Professionals

By : Rob VandenBrink
1 (1)
Book Image

Linux for Networking Professionals

1 (1)
By: Rob VandenBrink

Overview of this book

As Linux continues to gain prominence, there has been a rise in network services being deployed on Linux for cost and flexibility reasons. If you are a networking professional or an infrastructure engineer involved with networks, extensive knowledge of Linux networking is a must. This book will guide you in building a strong foundation of Linux networking concepts. The book begins by covering various major distributions, how to pick the right distro, and basic Linux network configurations. You'll then move on to Linux network diagnostics, setting up a Linux firewall, and using Linux as a host for network services. You'll discover a wide range of network services, why they're important, and how to configure them in an enterprise environment. Finally, as you work with the example builds in this Linux book, you'll learn to configure various services to defend against common attacks. As you advance to the final chapters, you’ll be well on your way towards building the underpinnings for an all-Linux datacenter. By the end of this book, you'll be able to not only configure common Linux network services confidently, but also use tried-and-tested methodologies for future Linux installations.
Table of Contents (20 chapters)
1
Section 1: Linux Basics
4
Section 2: Linux as a Network Node and Troubleshooting Platform
8
Section 3: Linux Network Services

Datacenter load balancer design considerations

Load balancing has been part of larger architectures for decades, which means that we've gone through several common designs.

The "legacy" design that we still frequently see is a single pair (or cluster) of physical load balancers that service all the load balanced workloads in the datacenter. Often, the same load balancer cluster is used for internal and external workloads, but sometimes, you'll see one internal pair of load balancers on the internal network, and one pair that only services DMZ workloads (that is, for external clients).

This model was a good approach in the days when we had physical servers, and load balancers were expensive pieces of hardware.

In a virtualized environment, though, the workload VMs are tied to the physical load balancers, which complicates the network configuration, limits disaster recovery options, and can often result in traffic making multiple "loops" between...