配置cluster,手動開啟的服務被cluster停止
各位,在虛擬機上做cluster的測試,資源和服務列表(cluster.conf)如下:
# more /etc/cluster/cluster.conf
<?xml version="1.0"?>
<cluster config_version="18" name="mycluster">
<fence_daemon post_fail_delay="0" post_join_delay="3"/>
<clusternodes>
<clusternode name="QT-ESA-rh5.4-A" nodeid="1" votes="1">
<fence>
<method name="1">
<device name="myfenced" nodename="QT-ESA-rh5.4-A"/>
</method>
</fence>
</clusternode>
<clusternode name="QT-ESA-rh5.4-B" nodeid="2" votes="1">
<fence>
<method name="1">
<device name="myfenced" nodename="QT-ESA-rh5.4-B"/>
</method>
</fence>
</clusternode>
</clusternodes>
<cman expected_votes="1" two_node="1"/>
<fencedevices>
<fencedevice agent="fence_manual" name="myfenced"/>
</fencedevices>
<rm>
<failoverdomains>
<failoverdomain name="myfailoverdomain" ordered="1" restricted="0">
<failoverdomainnode name="QT-ESA-rh5.4-B" priority="2"/>
<failoverdomainnode name="QT-ESA-rh5.4-A" priority="1"/>
</failoverdomain>
</failoverdomains>
<resources>
<ip address="192.168.122.13" monitor_link="1"/>
<netfs export="/sharedata" force_unmount="1" fstype="nfs" host="192.168.122.15" mountpoint="/opt/ultrapower/postgresql/data"
name="newnfs" options=""/>
<postgres-8 config_file="/opt/ultrapower/postgresql/data/postgresql.conf" name="mypost" postmaster_options="-D /opt/ultrapowe
r/postgresql/data" postmaster_user="postgres" shutdown_wait="30"/>
</resources>
<service autostart="1" domain="myfailoverdomain" name="srv_post" recovery="relocate">
<ip ref="192.168.122.13">
<netfs ref="newnfs">
<postgres-8 ref="mypost"/>
</netfs>
</ip>
</service>
</rm>
</cluster>
#
配置好后,在主節點(192.168.122.11)上手動開啟nfs和postgresql服務,然後開啟cluster服務(cman,rgmanager),結果cluster卻將在主節點上手動開啟的postgesql服務關閉了,/var/log/messages信息如下:
# tail -n 50 /var/log/messages
May 15 22:38:17 qt-esa-rh5 openais: New Configuration:
May 15 22:38:17 qt-esa-rh5 openais: r(0) ip(192.168.122.11)
May 15 22:38:17 qt-esa-rh5 openais: Members Left:
May 15 22:38:17 qt-esa-rh5 openais: Members Joined:
May 15 22:38:17 qt-esa-rh5 openais: CLM CONFIGURATION CHANGE
May 15 22:38:17 qt-esa-rh5 openais: New Configuration:
May 15 22:38:17 qt-esa-rh5 openais: r(0) ip(192.168.122.11)
May 15 22:38:17 qt-esa-rh5 openais: r(0) ip(192.168.122.12)
May 15 22:38:17 qt-esa-rh5 openais: Members Left:
May 15 22:38:17 qt-esa-rh5 openais: Members Joined:
May 15 22:38:17 qt-esa-rh5 openais: r(0) ip(192.168.122.12)
May 15 22:38:17 qt-esa-rh5 openais: This node is within the primary component and will provide service.
May 15 22:38:17 qt-esa-rh5 openais: entering OPERATIONAL state.
May 15 22:38:17 qt-esa-rh5 openais: got nodejoin message 192.168.122.11
May 15 22:38:17 qt-esa-rh5 openais: got nodejoin message 192.168.122.12
May 15 22:38:17 qt-esa-rh5 openais: got joinlist message from node 2
May 15 22:38:17 qt-esa-rh5 ccsd: Cluster is not quorate. Refusing connection.
May 15 22:38:17 qt-esa-rh5 ccsd: Error while processing connect: Connection refused
May 15 22:38:17 qt-esa-rh5 ccsd: Cluster is not quorate. Refusing connection.
May 15 22:38:17 qt-esa-rh5 ccsd: Error while processing connect: Connection refused
May 15 22:38:17 qt-esa-rh5 ccsd: Cluster is not quorate. Refusing connection.
May 15 22:38:17 qt-esa-rh5 ccsd: Error while processing connect: Connection refused
May 15 22:38:17 qt-esa-rh5 ccsd: Initial status:: Quorate
May 15 22:38:34 qt-esa-rh5 kernel: dlm: Using TCP for communications
May 15 22:38:35 qt-esa-rh5 kernel: dlm: got connection from 2
May 15 22:38:36 qt-esa-rh5 clurgmgrd: <notice> Resource Group Manager Starting
May 15 22:38:37 qt-esa-rh5 clurgmgrd: : <notice> Forcefully unmounting /opt/ultrapower/postgresql/data
May 15 22:38:38 qt-esa-rh5 clurgmgrd: : <warning> killing process 5059 (root bash /opt/ultrapower/postgresql/data)
May 15 22:38:38 qt-esa-rh5 clurgmgrd: : <warning> killing process 8888 (root bash /opt/ultrapower/postgresql/data)
May 15 22:38:38 qt-esa-rh5 clurgmgrd: : <warning> killing process 9820 (postgres bash /opt/ultrapower/postgresql/data)
May 15 22:38:38 qt-esa-rh5 clurgmgrd: : <warning> killing process 9856 (postgres postgres /opt/ultrapower/postgresql/data)
May 15 22:38:38 qt-esa-rh5 clurgmgrd: : <warning> killing process 9858 (postgres postgres /opt/ultrapower/postgresql/data)
May 15 22:38:38 qt-esa-rh5 clurgmgrd: : <warning> killing process 9859 (postgres postgres /opt/ultrapower/postgresql/data)
May 15 22:38:38 qt-esa-rh5 clurgmgrd: : <warning> killing process 9819 (root su /opt/ultrapower/postgresql/data)
May 15 22:38:43 qt-esa-rh5 avahi-daemon: Invalid legacy unicast query packet.
May 15 22:38:43 qt-esa-rh5 avahi-daemon: Received response from host 192.168.122.1 with invalid source port 1394 on interface 'eth0.0'
May 15 22:38:43 qt-esa-rh5 avahi-daemon: Invalid legacy unicast query packet.
May 15 22:38:43 qt-esa-rh5 avahi-daemon: Invalid legacy unicast query packet.
May 15 22:38:43 qt-esa-rh5 avahi-daemon: Received response from host 192.168.122.1 with invalid source port 1394 on interface 'eth0.0'
May 15 22:38:44 qt-esa-rh5 last message repeated 2 times
May 15 22:38:54 qt-esa-rh5 clurgmgrd: <notice> Starting stopped service service:srv_post
May 15 22:38:55 qt-esa-rh5 avahi-daemon: Registering new address record for 192.168.122.13 on eth0.
May 15 22:38:55 qt-esa-rh5 avahi-daemon: Invalid legacy unicast query packet.
May 15 22:38:55 qt-esa-rh5 avahi-daemon: Received response from host 192.168.122.1 with invalid source port 1396 on interface 'eth0.0'
May 15 22:38:55 qt-esa-rh5 avahi-daemon: Invalid legacy unicast query packet.
May 15 22:38:56 qt-esa-rh5 avahi-daemon: Invalid legacy unicast query packet.
May 15 22:38:56 qt-esa-rh5 avahi-daemon: Received response from host 192.168.122.1 with invalid source port 1396 on interface 'eth0.0'
May 15 22:38:56 qt-esa-rh5 avahi-daemon: Received response from host 192.168.122.1 with invalid source port 1396 on interface 'eth0.0'
May 15 22:38:57 qt-esa-rh5 clurgmgrd: <notice> Service service:srv_post started
May 15 22:38:57 qt-esa-rh5 avahi-daemon: Received response from host 192.168.122.1 with invalid source port 1396 on interface 'eth0.0'
#
好像是說從網關上收到了什麼無效的包,從而引起服務關閉。請各位大大給指點一下。
《解決方案》
如果你使用rgmanager,就不要手動啟服務。rgmanager啟動的時候會先stop服務,然後再start
《解決方案》
另外,你可以使用fence_xvm代替manual fence。
《解決方案》
能詳細介紹一下fencd_xvm是什麼東東嗎?對這個東西不太熟悉,謝謝,另:現在測試過程中出現新的提示如下:
May 16 12:49:15 qt-esa-rh5 openais: entering OPERATIONAL state.
May 16 12:49:15 qt-esa-rh5 openais: got nodejoin message 192.168.122.11
May 16 12:49:15 qt-esa-rh5 openais: got nodejoin message 192.168.122.12
May 16 12:51:38 qt-esa-rh5 kernel: dlm: Using TCP for communications
May 16 12:51:38 qt-esa-rh5 kernel: dlm: got connection from 2
May 16 12:51:38 qt-esa-rh5 clurgmgrd: <notice> Resource Group Manager Starting
May 16 12:51:40 qt-esa-rh5 clurgmgrd: : <err> Checking Existence Of File /opt/ultrapower/postgresql/data/postgresql.conf > Failed - File
Is Not Readable
May 16 12:51:40 qt-esa-rh5 clurgmgrd: : <err> Verifying Configuration Of postgres-8:mypost > Failed
May 16 12:51:40 qt-esa-rh5 clurgmgrd: <notice> stop on postgres-8 "mypost" returned 2 (invalid argument(s))
May 16 12:51:48 qt-esa-rh5 clurgmgrd: <notice> Starting stopped service service:srv_post
May 16 12:51:48 qt-esa-rh5 avahi-daemon: Registering new address record for 192.168.122.13 on eth0.
May 16 12:51:50 qt-esa-rh5 kernel: FS-Cache: Loaded
May 16 12:51:51 qt-esa-rh5 clurgmgrd: <notice> Service service:srv_post started
#