[ULTRAMONKEY-USERS] load balancing using ldirectord

[Glen Kendell]

If you use the direct routing method, your real servers need to have the
virtual IP as a secondary address with ARP suppressed.  Also make sure the
default gateway is the router and not the director.


On 1/30/08 12:35 AM, "Jordi Garcia" <jgarcia at pricewin.net> wrote:

> Hello,
> 
> I'm trying to configure a HA+LB environment with heartbeat+ldirectord.
> I followed this doc:
> http://www.ultramonkey.org/3/topologies/hc-ha-lb-eg.html
> 
> heartbeat 2.0.7-2
> ldirectord 1.2.5-3
> kernel 2.6.18-5-686
> 
> ha.cf
> ==========================================================
> logfacility local0
> use_logd yes
> 
> keepalive 1
> deadtime 10
> warntime 5
> initdead 150
> udpport 695
> bcast eth1
> auto_failback off
> node node1
> node node2
> respawn hacluster /usr/lib/heartbeat/ipfail
> apiauth ipfail gid=haclient uid=hacluster
> ==========================================================
> 
> haresources
> ==========================================================
> node1 ldirectord  LVSSyncDaemonSwap::master \
>       IPaddr2::10.2.1.200/23/eth1/10.2.1.255
> ==========================================================
> 
> ldirector.cf
> ==========================================================
> checktimeout=10
> checkinterval=2
> autoreload=no
> logfile="local0"
> quiescent=yes
> logfile="/var/log/ldirectord.log"
> 
> virtual=10.2.1.200:20057
>         real=10.2.1.20:20057 gate
>         real=10.2.1.109:20057 gate
>         real=10.2.1.110:20057 gate
>         service=none
>         scheduler=wrr
>         checktype=connect
>         protocol=tcp
> ==========================================================
> 
> I can see LVS kernel table without problems:
> 
> node1:# ipvsadm -L -n
> IP Virtual Server version 1.2.1 (size=4096)
> Prot LocalAddress:Port Scheduler Flags
>   -> RemoteAddress:Port           Forward Weight ActiveConn InActConn
> TCP  10.2.1.200:20057 wrr
>   -> 10.2.1.110:20057             Route   1      0          0
>   -> 10.2.1.109:20057             Route   1      0          0
>   -> 10.2.1.20:20057              Route   1      0          0
> 
> and VIP interface:
> 
> node1:/var/log# ip addr sh eth1
> 3: eth1: <BROADCAST,MULTICAST,UP,10000> mtu 1500 qdisc pfifo_fast qlen
> 1000
>     link/ether 00:30:48:41:9f:df brd ff:ff:ff:ff:ff:ff
>     inet 10.2.1.117/23 brd 10.2.1.255 scope global eth1
>     inet 10.2.1.200/23 brd 10.2.1.255 scope global secondary eth1
>     inet6 fe80::230:48ff:fe41:9fdf/64 scope link
>        valid_lft forever preferred_lft forever
> 
> 
> Well, the problem is that the service only works a few seconds:
> 
> client1:~# telnet 10.2.1.200 20057
> Trying 10.2.1.200...
> Connected to 10.2.1.200 (10.2.1.200).
> Escape character is '^]'.
> +OK Server Health 2.0.c
> quit
> Connection closed by foreign host.
> 
> Connection closed by foreign host.
> client1:~# telnet 10.2.1.200 20057
> Trying 10.2.1.200...
> Connected to 10.2.1.200 (10.2.1.200).
> Escape character is '^]'.
> +OK Server Health 2.0.c
> quit
> Connection closed by foreign host.
> 
> client1:~# telnet 10.2.1.200 20057
> Trying 10.2.1.200...
> telnet: connect to address 10.2.1.200: Connection refused
> telnet: Unable to connect to remote host: Connection refused
> 
> client1:~# telnet 10.2.1.200 20057
> Trying 10.2.1.200...
> telnet: connect to address 10.2.1.200: Connection refused
> telnet: Unable to connect to remote host: Connection refused
> 
> but, some seconds after...
> 
> client1:~# telnet 10.2.1.200 20057
> Trying 10.2.1.200...
> Connected to 10.2.1.200 (10.2.1.200).
> Escape character is '^]'.
> +OK Server Health 2.0.c
> quit
> Connection closed by foreign host.
> 
> [...]
> 
> some minutes after, the same situation.
> 
> 
> but LVS kernel table says:
> 
> node1:/var/log# ipvsadm -L -n
> IP Virtual Server version 1.2.1 (size=4096)
> Prot LocalAddress:Port Scheduler Flags
>   -> RemoteAddress:Port           Forward Weight ActiveConn InActConn
> TCP  10.2.1.200:20057 wrr
>   -> 10.2.1.110:20057             Route   1      0          2
>   -> 10.2.1.109:20057             Route   1      0          3
>   -> 10.2.1.20:20057              Route   1      0          2
> 
> 
> 
> Any ideas? Is it a problem with a real server? but If I try a telnet to
> 20057 port from master to all real servers, works fine..
> 
> 
> 
> Kind regards
> Jordi
> 
> _______________________________________________
> Ultramonkey-users mailing list
> Ultramonkey-users at vergenet.net
> http://lists.vergenet.net/listinfo/ultramonkey-users