haproxy

haproxy

HAProxy is a free, very fast and reliable solution offering high availability, load balancing, and proxying for TCP and HTTP-based applications

Source code can be found here

How to use it

# add emagtechlabs repository
helm repo add emagtechlabs  https://emagtechlabs.github.io/helm-charts/

# install haproxy chart
helm install emagtechlabs/haproxy

Prerequisites

• Helm v3
• Kubernetes 1.12+
• haproxy 2.1

Key	Type	Default	Description
annotations	object	{}	annotations
affinity	object	{}
configFiles	object	{"global.cfg": {}, "defaults.cfg": {}, "metrics.cfg": {}}	List of haproxy config files
checksumConfigMap.enabled	bool	true	doc
dnsPolicy	string	""	docs
dnsConfig	object	{}	docs
extraVolumes	list	[]	volumes
extraVolumeMounts	list	[]
fullnameOverride	string	""
image.repository	string	haproxy	docker image to be used
image.tag	string	2.2-alpine	docker image tag to be used
image.pullPolicy	string	IfNotPresent	Image pull policy
imagePullSecrets	list	[]
ingress.annotations."kubernetes.io/ingress.class"	string	"nginx"
ingress.enabled	bool	false	Enable ingress for stats interface
ingress.host	string	"host.example.com"
ingress.path	string	"/"
ingress.port	string	"metrics"	usually stats port should be exposed
ingress.tls	list	[]
livenessProbe	object	{}
nameOverride	string	""
nodeSelector	object	{}
podAnnotations	object	{}
podSecurityContext	object	{}
ports	object	{"metrics":{"containerPort":8404,"protocol":"TCP"}}	Every port that should be exposed to others apps should be listed here
readinessProbe	object	{}
replicaCount	int	1	Number of pods to be deployed
resources	object	{}	resource limits and request for container
securityContext	object	{}
service.annotations	object	{}
service.type	string	"ClusterIP"
service.clusterIP	string	""
service.loadBalancerIP	string	""
service.loadBalancerSourceRanges	list	[]
serviceAccount.annotations	object	{}
serviceAccount.create	bool	true
serviceMonitor.additionalLabels	object	{}
serviceMonitor.enabled	bool	false	If the operator is installed in your cluster, set to true to create a ServiceMonitor
serviceMonitor.interval	string	"15s"
serviceMonitor.metricsPath	string	"/metrics"	the path where metrics can be found
serviceMonitor.metricsPort	string	"metrics"	metrics port defined in ports section. Can be name of port or port itself
serviceMonitor.podTargetLabels	list	[]
serviceMonitor.targetLabels	list	[]
tolerations	list	[]

Default configuration

Default configuration for haproxy:

configFiles:
  global.cfg: |-
    global
      maxconn   10000
      log       127.0.0.1 local0
      nbproc    1
      nbthread  1

  defaults.cfg: |-
    defaults
      backlog   2000
      timeout   connect 10s
      timeout   client 30s
      timeout   server 30s
      timeout   check 10s
      log       global
      mode      tcp
      option    httplog
      maxconn   3000

  metrics.cfg: |-
    frontend metrics
      mode          http
      log           global
      maxconn       10
      bind          *:8404
      option        http-use-htx
      http-request  use-service prometheus-exporter if { path /metrics }
      stats         enable
      stats         uri /
      stats         refresh 10s
      stats         admin if LOCALHOST