A=MTBF/(MTBF+MTTR)

Design failure

Random failure

infant Mortality

wear out

User better components

Preemptively replace hardware prior to wear out

Peer review of all code

Simple design

Compact code foot print

heartbeat:

RHEL6.X   RHCS:corosync

RHEL5.X   RHCS:openais,cman,rgmanager

Corosync执行高可用应用程序的通信系统

corosync:Messaging layer

openais:

www.corosync.org

Diagnostics and failure analysis

corosync

Ha-aware

  crm(pacemaker)

corosync/heartbeat V3)

hawk

corosync-->pacemaker

  SUSE Linux Enterprise Server:hawk,webGUI

  LCMC:linux Cluster management Console

  RHCS:conga(luci/ricci)

    webGUI

  keepalived:VRRP,2节点

 

rpm,sources

resouce-agents

  pacemaker,corosync

            heartbeat

ldirectord

cluster-glue

pcs:

corosync:

  1、时间同步

  2、主机名

  3、SSH

ssh 172.16.100.6

#date

#ntpdate 172.16.0.1

#date

#ssh node1 'date'

#clear

#lftp 172.16.0.1/pub

#cd Sources/corosync/

#ls

#mget cluster-glue-* corosync-1.2.7-1.1.el5.i386.rpm

#mv openailslib-1.1.3-1.6.el5.i386.rpm /tmp

#ls

#scp *.rpm node1:/root

#ls /etc/yum.repos.d/

#wget ftp://172.16.0.1/pub/gls/server.repo -0 /etc/yum.repos.d/server.repo

#yum --nogpgcheck localinstall *.rpm

#yum -y --nogpgcheck localinstall *.rpm

#rpm -ql corosync

#cd /etc/corosync/

#ls

#cp corosync.conf.example corosync.conf

threads

fileline

#corosync-keygen

#ll

#file authkey

#scp -p authkey corosync.conf node2:/etc/corosync/

#mkdir /var/log/cluster

#ssh node2 'mkdir /var/log/cluster'

#service corosync start

#ssh node2 '/etc/init.d/corosync start'

#grep -e "Corosync Cluster Engine" -e  "configuration file" /var/log/cluster/corosync.log

 

#grep TOTEM /var/log/cluster/corosync.log

#grep ERROR: /var/log/cluter/corosync.log

#grep pcmk_startup  /var/log/cluster/corosync.log

#crm_mon