1

# Design and Implementation of the <a href="https://github.com/qmsk/clusterf"><tt>clusterf</tt></a> Load Balancer for Docker Clusters

     2

     3

### Author

     4

Tero Marttila

     5

     6

### Submitted

     7

Thesis submitted for examination for the degree of Master of Science in Technology at <a href="http://comnet.aalto.fi/">Aalto University Department of Communications and Networking</a>,

     8

     9

Espoo 10.10.2016

    10

    11

### Thesis supervisor

    12

Prof. Raimo Kantola

    13

    14

### Thesis advisor

    15

PhD. Pasi Sarolahti

    16

    17

## Seminar Presentation

    18

    19

Presented 08.10.2016 ([slides](https://mycourses.aalto.fi/pluginfile.php/292054/mod_folder/content/0/Marttila_071016.pdf?forcedownload=1)].

    20

    21

## Abstract

    22

    23

### Relation to a bigger picture

    24

[Docker](https://www.docker.com/) uses the Linux container `namespace` and `cgroup` primitives to provide isolated application runtime environments, allowing the distribution of self-contained application images that can be run in the form of Docker containers.

    25

    26

Container platforms provide the infrastructure needed to deploy horizontally scalable container services into the Cloud, including orchestration, networking, service discovery and load balancing.

    27

    28

### Purpose

    29

This thesis studies container networking architectures and existing Cloud load balancer implementations to create a design for a scalable load balancer using the [Linux IPVS](http://www.linuxvirtualserver.org/software/ipvs.html) network-level load balancer implementation.

    30

    31

The `clusterf` load balancer architecture uses a two-level load balancing scheme combining different packet forwarding methods for scalability and           compatibility with existing Docker applications.

    32

    33

A distributed load balancer control plane is implemented to provide automatic load balancing for Docker containers.

    34

    35

### Research method

    36

The `clusterf` load balancer is evaluated using a testbed environment, measuring the performance the network-level Linux IPVS implementation.

    37

    38

The scalability of the `clusterf` load balancer is tested using Equal-Cost Multi-Path (ECMP) routing and [IPVS connection synchronization](http://www.linux-vs.org/docs/sync.html).

    39

    40

### Key results

    41

The result is that the network-level Linux IPVS load balancer performs significantly better than the application-level HAProxy load balancer in the same      configuration.

    42

    43

The `clusterf` design allows for horizontal scaling with connection failover between IPVS load balancers.

    44

    45

### Practical implications

    46

The current `clusterf` implementation requires the use of asymmetric routing within a network, such as provided by local Ethernet networks.

    47

    48

Extending the `clusterf` design to support deployment onto existing Cloud infrastructure platforms with different networking implementations would qualify   the `clusterf` load balancer for use in container platforms.

    49

    50

## Thesis

    51

    52

Pending publication.

    53

    55

    56

## [The `clusterf` Load Balancer](https://github.com/qmsk/clusterf)

    57

    58

### Design

    59

The architecture of the `clusterf` load balancer is shown in the Figure below.

    60

    61

<img src="//static.qmsk.net/www/masters/clusterf-architecture.svg" style="width: 100%" />

    62

    63

### Implementation

    64

The Go implementation of the `clusterf` load balancer control plane is available on GitHub: 

    65

    66

> [`go get github.com/qmsk/clusterf`](https://github.com/qmsk/clusterf)