HAProxy Example
HAProxy is a free load balancer that runs in Linux. It is very powerful and supports monitoring capabilities out of the box.
For information about its configuration, see:
Global Settings
Global settings are applied across frontend and backend servers and can generally be ignored.
For example:
#--------------------------------------------------------------------- #
Global settings
#---------------------------------------------------------------------
global
# setup logging and force level to debug. Logs require rsyslog be setup.
chroot /var/lib/haproxy
log /dev/log local0 debug
# maximum number of connections allowed
maxconn 10000
# turn on stats unix socket
stats socket /run/haproxy/admin.sock mode 660 level admin expose-fd listeners
stats timeout 30s
# user and group haproxy will run as
user haproxy
group haproxy
Default Settings
These settings are for default configurations of the backend servers.
• log
Use the logging configured in Global Settings.
• mode
Determines how requests are parsed. In this example, we use http handling.
• option
For example:
#---------------------------------------------------------------------
# common defaults that all the 'listen' and 'backend' sections will
# use if not designated in their block
#---------------------------------------------------------------------
defaults
default-server init-addr last,libc,none
log global
mode http
option httplog
option redispatch
option forwardfor
timeout connect 10s
timeout client 60s
timeout server 60s
timeout tunnel 1h
FrontEnd Settings
This is the frontend through which all applications will go. It contains the routing rules that determine which backend server is used to handle the incoming request.
• bind
Indicates the port on which incoming requests will be received.
• capture
Includes the session ID cookie in the HAProxy http request logging.
• acl
Creates a variable based on the condition being checked. For example, -i -m beg <string> checks the URI beginning with <string>.
• use_backend
Based on the true or false value of variables, routs to a specific group of backend servers.
• default
The backend used if it doesn’t match the other rules.
For example:
#---------------------------------------------------------------------
# FrontEnd Configuration
#---------------------------------------------------------------------
frontend ft_web
bind *:8080 name web
# log the session cookie if passed
capture cookie JSESSIONID= len 32
## path based routing to connection server
acl path_cxserver path -i -m beg /Thingworx/WS
acl path_tunnelserver path -i -m beg /Thingworx/WSTunnelServer
acl path_tunnelclient path -i -m beg /Thingworx/WSTunnelClient
use_backend cxserver if path_cxserver or path_tunnelserver or path_tunnelclient
# default traffic to platform
default_backend platform
Example for ThingWorx Flow
#---------------------------------------------------------------------
# FrontEnd Configuration
#---------------------------------------------------------------------
frontend ft_web
bind *:8443 ssl crt /certs/haproxy.pem name sslweb
# log the session cookie if passed
capture cookie JSESSIONID= len 32
##path based routing to ThingWorx Flow
acl p_flow1 path -i -m beg /Thingworx/Composer/apps/flow
acl p_flow2 path -i -m beg /Thingworx/Flow
acl p_flow3 path -i -m beg /Thingworx/Triggers
acl p_flow4 path -i -m beg /Thingworx/Lookups
acl p_flow5 path -i -m beg /Thingworx/Oauths
acl p_flow6 path -i -m beg /Thingworx/Subsystems/EventProcessingSubsystem/Subscriptions
acl p_flow7 path -i -m beg /enterprise/v1/fetchconfig
use_backend flow if p_flow1 or p_flow2 or p_flow3 or p_flow4 or p_flow5 or p_flow6 or p_flow7
BackEnd Platform Settings
The backend is a group of servers that can handle requests passed from the frontend and any specific configurations.
• balance
The load balancing algorithm. The default is round robin, which walks through available servers for the next available connection.
• cookie
Inserts a cookie used by the load balancer to route future requests to the same server.
For example:
#---------------------------------------------------------------------
# BackEnd Platform Configuration
#---------------------------------------------------------------------
backend platform
balance roundrobin
# sticky sessions
cookie SERVER insert indirect nocache
http-request set-header X-Forwarded-Port %[dst_port]
http-request add-header X-Forwarded-Proto https if { ssl_fc }
# health check
option httpchk GET /Thingworx/health
# configure platform instances
server platform1 10.0.10.10:8080 check inter 1000 fastinter 1000 cookie twx1
server platform2 10.0.10.11:8080 check inter 1000 fastinter 1000 cookie twx2
BackEnd Connection Server Settings
Example:
#---------------------------------------------------------------------
# BackEnd Connection Server Configuration
#---------------------------------------------------------------------
backend cxserver
balance source
#hash-type consistent
option httpchk GET /
# configure connection server instances
server cxserver1 10.0.10.20:8080 check port 9009
server cxserver2 10.0.10.21:8080 check port 9009
Monitoring
HAProxy has a monitoring UI that can be used to view traffic flow. You can configure the incoming port and optional credentials.
#---------------------------------------------------------------------
#HAProxy Monitoring Config
#---------------------------------------------------------------------
listen stats
bind *:1936
mode http
option forwardfor
option httpclose
stats enable
stats uri /
stats refresh 5s
stats show-legends
stats realm Haproxy\ Statistics
Full Example
#---------------------------------------------------------------------
# Global settings
#---------------------------------------------------------------------
global
# setup logging and force level to debug. Logs require rsyslog be setup.
chroot /var/lib/haproxy
log /dev/log local0 debug
# maximum number of connections allowed
maxconn 10000
# turn on stats unix socket
stats socket /run/haproxy/admin.sock mode 660 level admin expose-fd listeners
stats timeout 30s
# user and group haproxy will run as
user haproxy
group haproxy
#---------------------------------------------------------------------
# common defaults that all the 'listen' and 'backend' sections will
# use if not designated in their block
#---------------------------------------------------------------------
defaults
default-server init-addr last,libc,none
log global
mode http
option httplog
option redispatch
option forwardfor
timeout connect 10s
timeout client 60s
timeout server 60s
timeout tunnel 1h
#---------------------------------------------------------------------
# FrontEnd Configuration
#---------------------------------------------------------------------
frontend ft_web
bind *:8080 name web
# log the session cookie if passed
capture cookie JSESSIONID= len 32
## path based routing to connection server
acl path_cxserver path -i -m beg /Thingworx/WS
acl path_tunnelserver path -i -m beg /Thingworx/WSTunnelServer
acl path_tunnelclient path -i -m beg /Thingworx/WSTunnelClient
use_backend cxserver if path_cxserver or path_tunnelserver or path_tunnelclient
# default traffic to platform
default_backend platform
#---------------------------------------------------------------------
# BackEnd Platform Configuration
#---------------------------------------------------------------------
backend platform
balance roundrobin
# sticky sessions
cookie SERVER insert indirect nocache
http-request set-header X-Forwarded-Port %[dst_port]
http-request add-header X-Forwarded-Proto https if { ssl_fc }
# health check
option httpchk GET /Thingworx/health
# configure platform instances
server platform1 10.0.10.10:8080 check inter 1000 fastinter 1000" cookie twx1
server platform2 10.0.10.11:8080 check inter 1000 fastinter 1000" cookie twx2
#---------------------------------------------------------------------
# BackEnd Connection Server Configuration
#---------------------------------------------------------------------
backend cxserver
balance source
#hash-type consistent
option httpchk GET /
# configure connection server instances
server cxserver1 10.0.10.20:8080 check port 9009
server cxserver2 10.0.10.21:8080 check port 9009
#---------------------------------------------------------------------
#HAProxy Monitoring Config
#---------------------------------------------------------------------
listen stats
bind *:1936
mode http
option forwardfor
option httpclose
stats enable
stats uri /
stats refresh 5s
stats show-legends
stats realm Haproxy\ Statistics
Configuring SSL/TLS for HAProxy
To configure SSL/TLS for HAProxy, see
Configuring SSL/TLS for HAProxy.