Skip to content

Latest commit

 

History

History
105 lines (95 loc) · 4.21 KB

README.md

File metadata and controls

105 lines (95 loc) · 4.21 KB
Raw

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