Lastverteilung mit HAProxy¶
Mit Varnish lässt sich zwar eine einfache Lastverteilung im Round-Robin-Verfahren realisieren, für zuverlässige Hochverfügbarkeit sind jedoch aufwendigere Verfahren notwendig.
Dies kann mit HAProxy realisiert werden. Dabei lässt sich HAProxy einfach mit Buildout installieren:
[haproxy-build]
recipe = plone.recipe.haproxy
url = http://www.haproxy.org/download/1.4/src/haproxy-1.4.26.tar.gz
target = linux26
cpu = i686
pcre = 1
[haproxy-config]
recipe = collective.recipe.template
input = ${buildout:directory}/templates/haproxy.conf.in
output = ${buildout:directory}/etc/haproxy.conf
user = haproxy
group = haproxy
frontend-host = 127.0.0.1
frontend-port = 8001
Die haproxy.conf.in-Datei in templates sieht dann z.B. so aus:
global
maxconn ${haproxy-conf:maxconn}
user ${haproxy-config:user}
group ${haproxy-config:group}
daemon
nbproc 1
spread-checks 3
ulimit-n 65536
defaults
mode http
# The zope instances have maxconn 1, and it is important that
# unused/idle connections are closed as soon as possible.
option httpclose
# Remove requests from the queue if people press stop button
option abortonclose
retries 3
option redispatch
monitor-uri /haproxy-ping
timeout connect 5s
timeout queue 30s
timeout client 50s
timeout server 120s
timeout check 50s
stats enable
stats uri /haproxy-status
stats refresh 60s
stats realm Haproxy\ statistics
frontend zeocluster
bind ${haproxy-config:frontend-host}:${haproxy-config:frontend-port}
default_backend zope
option httplog
log 127.0.0.1:1514 local6
# Load balancing over the zope instances
backend zope
# Use Zope's __ac cookie as a basis for session stickiness if present.
appsession __ac len 32 timeout 1d
# Otherwise add a cookie called "serverid" for maintaining session stickiness.
# This cookie lasts until the client's browser closes, and is invisible to Zope.
cookie serverid insert nocache indirect
# If no session found, use the roundrobin load-balancing algorithm to pick a backend.
balance roundrobin
# Use / (the default) for periodic backend health checks
option httpchk GET /
# Server options:
# "maxconn" is how many connections can be sent to the server at once
# "check" enables health checks
# "rise 1" means consider Zope up after 1 successful health check
server instance1 127.0.0.1:${instance1:http-address} weight 1 check inter 15s rise 2 fall 1 maxconn 2
server instance2 127.0.0.1:${instance1:http-address} weight 1 check inter 15s rise 2 fall 1 maxconn 2
HAProxy kann dann gestartet werden mit:
${buildout:directory}/bin/haproxy -f ${buildout:directory}/etc/haproxy.conf -db
Unter http://localhost:8001/haproxy-status können Sie sich dann den aktuellen Status des HAProxy anschauen:
Siehe auch