HAproxy Use backup server with active servers in one backend

haproxy

I'm struggling with a scenario that I've searched and read so much but couldn't find anything to solve it.
I have 3 servers, 2 actives and one as a standby(backup). every time one of active servers go down, the standby server will be active and traffic load balance between 2 server (not just standby server).
I am able to solve this problem with two backends, BUT I need just one backend.

#### Fronend Configuration ##### 
frontend webserver 

bind 192.168.64.100:80
default_backend appserver


#### Backend Configuration #####
backend appserver 
balance roundrobin
server Node-1 192.168.64.132:80 check
server Node-2 192.168.64.133:80 check

server Node-3 192.168.64.134:80 check

I want Node-3 be the standby server. if Node-1 or Node-2 isn't accessible, Node-3 take the place

Best Answer

Take a look here: https://www.haproxy.com/blog/failover-and-worst-case-management-with-haproxy/

It looks like by default the backup servers only take effect when all of the primary servers are offline. So it seems like if Node-1 is down and Node-2 is still up, it should still be serving with only Node-2. If both Node-1 and Node-2 are offline, Node-3 would be used. It's possible your situation is occurring because some config is indicating that when 50% of servers are unavailable, the backup should be triggered (perhaps this is default?).

But if you look further at that page, you can set logic to trigger a backup group when a certain amount of servers in the primary fail. With only three servers in the set, I'm not sure how you'd properly accomplish that bilaterally but you could set if 1 of the 2 fails, trigger a backup of two which includes a primary. Honestly it make more sense to leverage all three in the primary group unless you pay a penalty for using it (e.g. hosted services).

frontent ft_app
 bind 10.0.0.1:80
# detect capacity issues in production farm
 acl MAIN_not_enough_capacity nb_srv(bk_app_main) le 2
# failover traffic to backup farm
 use_backend bk_app_backup if MAIN_not_enough_capacity
 default_backend bk_app_main
backend bk_app_main
 server s11 10.0.0.101:80 check
 server s12 10.0.0.102:80 check
 server s13 10.0.0.103:80 check
 server s14 10.0.0.104:80 check
backend bk_app_backup
 option allbackups
 server s21 20.0.0.101:80 check
 server s22 20.0.0.102:80 check
 server s23 20.0.0.103:80 check backup
 server s24 20.0.0.104:80 check backup

Related Solutions

Linux – HAProxy with ‘backup’ failure

Addressing the original question as to why the example does not work there are 2 things which stand out:

the check keyword is not added to the server declarations, which tells haproxy to use the health check directive on this server.
the httpchk keyword is best used together with a port on the server directive to define where the HTTP health check service is running. This method assumes you've setup a service somewhere that reports the health of the master.

For example, I think something like this would work (untested)

backend mysql
  mode tcp
  balance roundrobin
  option  httpchk GET /master_status
  server db01 10.0.0.236:3306 weight 100 check port 3305 inter 1s rise 3 fall 2
  server db02 10.0.0.68:3306 weight 100 check port 3305 inter 1s rise 3 fall 2 backup

In this example, haproxy will query the /master_status url on port 3305 on each server. If the server returns a 2xx status code, the backend will be available through haproxy. Otherwise, it will be offline.

To solve the problem of the health check service, you can use the rubygem mysql_health. You would run this service on each database server that participates in the haproxy cluster.

https://github.com/osterman/mysql_health

To start the health check service, you would run something like:

/usr/bin/mysql_health --server:pid-file=/var/run/mysql_health.pid \
                      --log:file /var/log/mysql_health.log \
                      --server:daemonize \
                      --server:port 3305 \
                      --check:master \
                      --check:database mysql

There are many more command line options available. Use the --help argument for more details.

Mysql – Pacemaker Active/Active haproxy load balancing

I have active-active 2*virtIP cluster

For CRM config:
I'm using two virtual IPs as primitive IPaddr2 services
and for the service which should run on both nodes:

create primitive for it, you'll then use it's ID
make "clone" from it like:
clone any_name_of_the_clone your_primitive_service_id \ meta clone-max="2" clone-node-max="1"

You may add order (to start virt IP after starting clone - NOT primitive, after you create clone you should not use it's child's ID)

It is working, failover works (assigning 2 IPs on one node when other fails).
However I have problem how to make colocation - I mean to have contrained services: I cannot have virtIP on node with subservice failed.
It is OK when the service is down - cluster brings it up but when starting fails (e.g. broken config for the service) - cluster notes error but brings IP up.

Anyone knows what's the cause?
- Is it a matter of bad monitoring/start/stop control or it is a matter of configuring constraints?

EDIT:
I have added to Primitive option to 'op start': on-fail="standby". Now when my service (the only primitive in clone) cannot start the node looses also virtIP
This seems to solve my problem now.

Best Answer

Related Solutions

Linux – HAProxy with ‘backup’ failure

Mysql – Pacemaker Active/Active haproxy load balancing

Related Topic