Впервые попробовал собрать двухнодовый кластер на базе двух vm OracleLinux на ESX. Кворумный диск не использую. Ноды: a-svfeOL 10.10.60.15 b-svfeOL 10.10.60.16 общий ip 10.10.06.14
Конфиг кластера:
<?xml version="1.0"?>
<cluster config_version="3" name="svfeOL-cluster">
<cman expected_votes="1" two_node="1"/>
<clusternodes>
<clusternode name="a-svfeOL" nodeid="1" votes="1">
<fence/>
</clusternode>
<clusternode name="b-svfeOL" nodeid="2" votes="1">
<fence/>
</clusternode>
</clusternodes>
<fencedevices/>
<rm log_facility="local4" log_level="5">
<failoverdomains>
<failoverdomain name="svfeOL-cluster" nofailback="0" ordered="0" restricted="1">
<failoverdomainnode name="a-svfeOL" priority="1"/>
<failoverdomainnode name="b-svfeOL" priority="1"/>
</failoverdomain>
</failoverdomains>
<service autostart="1" name="svfeOL" recovery="relocate">
<ip address="10.10.60.14" monitor_link="1"/>
</service>
</rm>
<totem consensus="4800" join="60" token="20000" token_retransmits_before_loss_const="20"/>
</cluster>
Cluster Status for svfeOL-cluster @ Fri Sep 21 12:05:34 2012
Member Status: Quorate
Member Name ID Status
------ ---- ---- ------
a-svfeOL 1 Online, rgmanager
b-svfeOL 2 Online, Local, rgmanager
Service Name Owner (Last) State
------- ---- ----- ------ -----
service:svfeOL a-svfeOL started
Вывод cman_tool status:
Version: 6.2.0
Config Version: 3
Cluster Name: svfeOL-cluster
Cluster Id: 4149
Cluster Member: Yes
Cluster Generation: 180
Membership state: Cluster-Member
Nodes: 2
Expected votes: 1
Total votes: 2
Node votes: 1
Quorum: 1
Active subsystems: 8
Flags: 2node
Ports Bound: 0 177
Node name: a-svfeOL
Node ID: 1
Multicast addresses: 239.192.16.69
Node addresses: 10.10.60.15
На второй почти тоже самое, только ID 2 и Ports Bound: пусто.
Для организации общего IP 10.10.60.14 был установлен hearbeat 3.0.3. hc.cf:
debugfile /var/log/ha-debug
logfile /var/log/ha-log
logfacility local0
keepalive 2
deadtime 25
warntime 10
initdead 50
udpport 694
bcast eth4
ucast eth4 10.10.60.15
auto_failback on
node a-svfeOL
node b-svfeOL
a-svfeOL IPaddr::10.10.60.14/24/eth4
Cluster Status for svfeOL-cluster @ Fri Sep 21 12:21:12 2012
Member Status: Quorate
Member Name ID Status
------ ---- ---- ------
a-svfeOL 1 Online, Local, rgmanager
b-svfeOL 2 Offline
Service Name Owner (Last) State
------- ---- ----- ------ -----
service:svfeOL a-svfeOL started
dlm: closing connection to node 2
dlm: closing connection to node 1
INFO: task rgmanager blocked for more than 120 secons
Unfencing self... fence_node: cannot connect to cman
Leaving fence domain... found dlm lockspace /sys/kernel/dlm/rgmanager