教学文章
Technology Exchange
400电话

免费咨询热线
400-090-9964

教学文章

Vmware 下Oracle RAC搬家

时间:2017-10-13 来源:

 最近虚拟机下的Oracle 10g RAC搬家,搬家完毕之后,Oracle 集群resource之VIP无法正常启动,收到了CRS-0233: Resource or relatives are currently involved with another operation 错误提示。为为啥呢,原来啊,搬家了地址发生变化了,你得使用你家里的新地址阿....

  1、环境描述

  Oracle 10g RAC + Suse 10

  注,将RAC虚拟机搬家之后,通常情况下我们在添加虚拟机时选择复制(移动这个没试过,不知道是否会有问题)

  其次,由于选择了复制,虚拟机要生成一个新的UUID(UUID是指在一台机器上生成的数字,它保证对在同一虚拟环境中的机器唯一性)。

  同时MAC地址以及网络接口名也会发生相应的变化(首次启动虚拟机原来的eth0以及eth1不可用),通常情况下需要对此修改。

  不同的Linux系统对新网络接口有不同的处理方法。Oracle Linux以及RedHat可以直接把原来的删除然后将新的网络接口名重命名为原来的。

  2、CRS-1006/CRS-0215/CRS-0233错误

  #修改网卡之后,重新启动两个节点

  #resource vip 状态为offline

  oracle@bo2dbp:~> ./crs_stat.sh |grep bo2dbp

  Resource name Target State

  -------------- ------ -----

  ora.bo2dbp.ASM1.asm ONLINE ONLINE on bo2dbp

  ora.bo2dbp.LISTENER_BO2DBP.lsnr ONLINE OFFLINE

  ora.bo2dbp.LISTENER_ORA10G_BO2DBP.lsnr ONLINE OFFLINE

  ora.bo2dbp.gsd ONLINE ONLINE on bo2dbp

  ora.bo2dbp.ons ONLINE OFFLINE

  ora.bo2dbp.vip ONLINE OFFLINE

  ora.ora10g.db ONLINE ONLINE on bo2dbp

  ora.ora10g.ora10g1.inst ONLINE ONLINE on bo2dbp

  #尝试手动启动ons

  oracle@bo2dbp:~> crs_start ora.bo2dbp.ons

  Attempting to start `ora.bo2dbp.ons` on member `bo2dbp`

  Start of `ora.bo2dbp.ons` on member `bo2dbp` failed.

  CRS-1006: No more members to consider

  CRS-0215: Could not start resource 'ora.bo2dbp.ons'.

  #通过onsctl方式启动也收到失败信息

  oracle@bo2dbp:~> onsctl start

  Number of onsconfiguration retrieved, numcfg = 2

  onscfg[0]

  {node = bo2dbp.2gotrade.com, port = 6200}

  Adding remote host bo2dbp.2gotrade.com:6200

  onscfg[1]

  {node = bo2dbs.2gotrade.com, port = 6200}

  Adding remote host bo2dbs.2gotrade.com:6200

  Number of onsconfiguration retrieved, numcfg = 2

  onscfg[0]

  {node = bo2dbp.2gotrade.com, port = 6200}

  Adding remote host bo2dbp.2gotrade.com:6200

  onscfg[1]

  {node = bo2dbs.2gotrade.com, port = 6200}

  Adding remote host bo2dbs.2gotrade.com:6200

  onsctl: ons failed to start

  #尝试手动启动vip,收到了CRS-0233错误

  oracle@bo2dbp:~> crs_start ora.bo2dbp.vip

  CRS-0233: Resource or relatives are currently involved with another operation.

  3、分析故障

  #查看节点bo2dbp主机上的ip地址

  oracle@bo2dbp:~> ifconfig #当前系统的网络接口名为eth2, eth5

  eth2 Link encap:Ethernet HWaddr 00:0C:29:4A:66:28

  inet addr:192.168.7.51 Bcast:192.168.7.255 Mask:255.255.255.0

  eth5 Link encap:Ethernet HWaddr 00:0C:29:4A:66:32

  inet addr:10.10.7.51 Bcast:10.10.7.255 Mask:255.255.255.0

  #查看节点bo2dbp上集群网络层的配置,网络接口名与实际的网卡名一致

  oracle@bo2dbp:~> oifcfg iflist

  eth2 192.168.7.0

  eth5 10.10.7.0

  oracle@bo2dbp:~> oifcfg getif -global #此处发现网络接口名与实际的网卡名不一致

  eth3 192.168.7.0 global public

  eth4 10.10.7.0 global cluster_interconnect

  #查看节点bo2dbs主机上的ip地址

  oracle@bo2dbs:~> ifconfig

  eth5 Link encap:Ethernet HWaddr 00:0C:29:27:43:EB

  inet addr:10.10.7.52 Bcast:10.10.7.255 Mask:255.255.255.0

  eth6 Link encap:Ethernet HWaddr 00:0C:29:27:43:E1

  inet addr:192.168.7.52 Bcast:192.168.7.255 Mask:255.255.255.0

  #查看节点bo2dbp上集群网络层的配置,网络接口名与实际的网卡名一致

  oracle@bo2dbs:~> oifcfg iflist

  eth6 192.168.7.0

  eth5 10.10.7.0

  oracle@bo2dbs:~> oifcfg getif -global #此处同样发现网络接口名与实际的网卡名不一致,应该是原来的网络接口名

  eth3 192.168.7.0 global public

  eth4 10.10.7.0 global cluster_interconnect

  #从上面的情况来看,各个节点的网络接口名不一致,同时网络集群层的配置信息使用了原来的接口名,应当对其更新

  #为了统一网络接口名,下面将其使用统一的名字为bond1, bond2

  4、解决故障

  #将网卡统一重命名,方法参照:http://blog.csdn.net/robinson_0612/article/details/8131771

  #下面是重命名后的结果

  oracle@bo2dbp:~> oifcfg iflist

  bond1 192.168.7.0

  bond2 10.10.7.0

  oracle@bo2dbs:~> oifcfg iflist

  bond1 192.168.7.0

  bond2 10.10.7.0

  #下面的查询中集群层的public与cluster_interconnect还是之前的配置信息

  #应该需要改成一致,此处我们先不改,看看会出现什么错误

  oracle@bo2dbp:~> oifcfg getif -global

  eth3 192.168.7.0 global public

  eth4 10.10.7.0 global cluster_interconnect

  oracle@bo2dbs:~> oifcfg getif -global

  eth3 192.168.7.0 global public

  eth4 10.10.7.0 global cluster_interconnect

  #重新启动crs

  oracle@bo2dbp:~> sudo -s /u01/oracle/crs/bin/crsctl start crs

  root'''s password:

  Attempting to start CRS stack

  The CRS stack will be started shortly

  #下面的查询表明crs后台进程正常

  oracle@bo2dbp:~> crsctl check crs

  CSS appears healthy

  CRS appears healthy

  EVM appears healthy

  #下面查询的结果还是跟以前一样

  oracle@bo2dbp:~> ./crs_stat.sh |grep bo2dbp

  Resource name Target State

  -------------- ------ -----

  ora.bo2dbp.ASM1.asm ONLINE ONLINE on bo2dbp

  ora.bo2dbp.LISTENER_BO2DBP.lsnr ONLINE OFFLINE

  ora.bo2dbp.LISTENER_ORA10G_BO2DBP.lsnr ONLINE OFFLINE

  ora.bo2dbp.gsd ONLINE ONLINE on bo2dbp

  ora.bo2dbp.ons ONLINE OFFLINE

  ora.bo2dbp.vip ONLINE OFFLINE

  ora.ora10g.db ONLINE ONLINE on bo2dbp

  ora.ora10g.ora10g1.inst ONLINE ONLINE on bo2dbp

  #将所有的资源全部关闭

  oracle@bo2dbp:~> crs_stop -all

  #使用oifcfg修改集群层网络配置

  oracle@bo2dbp:~> oifcfg delif -global

  oracle@bo2dbp:~> oifcfg getif -global

  oracle@bo2dbp:~> oifcfg setif -global bond1/192.168.7.0:public

  oracle@bo2dbp:~> oifcfg setif -global bond2/10.10.7.0:cluster_interconnect

  oracle@bo2dbp:~> oifcfg getif -global

  bond1 192.168.7.0 global public

  bond2 10.10.7.0 global cluster_interconnect

  #reboot之后,资源状态还是跟之前一样

  #先查看vip日志信息,我们先来解决vip的问题

  bo2dbp:/u01/oracle/crs/log/bo2dbp/racg # tail -50 ora.bo2dbp.vip.log

  2012-12-28 11:25:13.783: [RACG][2151948784] [16581][2151948784][ora.bo2dbp.vip]:

  clsrcexecut: env ORACLE_CONFIG_HOME=/u01/oracle/crs

  2012-12-28 11:25:13.783: [RACG][2151948784] [16581][2151948784][ora.bo2dbp.vip]:

  clsrcexecut: cmd = /u01/oracle/crs/bin/racgeut -e _USR_ORA_DEBUG=0 54 /u01/oracle/crs/bin/racgvip start bo2dbp

  2012-12-28 11:25:13.783: [RACG][2151948784] [16581][2151948784][ora.bo2dbp.vip]: clsrcexecut: rc = 1, time = 3.220s

  2012-12-28 11:25:16.979: [RACG][2151948784] [16581][2151948784][ora.bo2dbp.vip]:

  clsrcexecut: env ORACLE_CONFIG_HOME=/u01/oracle/crs

  2012-12-28 11:25:16.979: [RACG][2151948784] [16581][2151948784][ora.bo2dbp.vip]: clsrcexecut:

  cmd = /u01/oracle/crs/bin/racgeut -e _USR_ORA_DEBUG=0 54 /u01/oracle/crs/bin/racgvip check bo2dbp

  2012-12-28 11:25:16.979: [RACG][2151948784] [16581][2151948784][ora.bo2dbp.vip]: clsrcexecut: rc = 1, time = 3.190s

  2012-12-28 11:25:16.979: [RACG][2151948784] [16581][2151948784][ora.bo2dbp.vip]: end for resource =

  ora.bo2dbp.vip, action = start, status = 1, time = 6.430s

  2012-12-28 11:25:23.807: [RACG][695611888] [17488][695611888][ora.bo2dbp.vip]: eth3: error

  fetching interface information: Device not found #此处提示eth3没有找到,我们希望使用的是bond1

  checkIf: interface eth3 is down

  Invalid parameters, or failed to bring up VIP (host=bo2dbp)

  2012-12-28 11:25:23.807: [RACG][695611888] [17488][695611888][ora.bo2dbp.vip]:

  clsrcexecut: env ORACLE_CONFIG_HOME=/u01/oracle/crs

  2012-12-28 11:25:23.807: [RACG][695611888] [17488][695611888][ora.bo2dbp.vip]: clsrcexecut:

  cmd = /u01/oracle/crs/bin/racgeut -e _USR_ORA_DEBUG=0 54 /u01/oracle/crs/bin/racgvip start bo2dbp

  2012-12-28 11:25:23.807: [RACG][695611888] [17488][695611888][ora.bo2dbp.vip]: clsrcexecut:

  rc = 1, time = 3.220s

  2012-12-28 11:25:27.018: [RACG][695611888] [17488][695611888][ora.bo2dbp.vip]: clsrcexecut:

  env ORACLE_CONFIG_HOME=/u01/oracle/crs

  2012-12-28 11:25:27.018: [RACG][695611888] [17488][695611888][ora.bo2dbp.vip]: clsrcexecut:

  cmd = /u01/oracle/crs/bin/racgeut -e _USR_ORA_DEBUG=0 54 /u01/oracle/crs/bin/racgvip check bo2dbp

  2012-12-28 11:25:27.018: [RACG][695611888] [17488][695611888][ora.bo2dbp.vip]: clsrcexecut: rc = 1, time = 3.210s

  2012-12-28 11:25:27.018: [RACG][695611888] [17488][695611888][ora.bo2dbp.vip]: end for resource =

  ora.bo2dbp.vip, action = start, status = 1, time = 6.450s

  2012-12-28 11:25:33.822: [RACG][3600347632] [18308][3600347632][ora.bo2dbp.vip]: eth3: error

  fetching interface information: Device not found #再次出现eth3没有找到的错误提示

  checkIf: interface eth3 is down

  Invalid parameters, or failed to bring up VIP (host=bo2dbp)

  2012-12-28 11:25:33.822: [RACG][3600347632] [18308][3600347632][ora.bo2dbp.vip]: clsrcexecut:

  env ORACLE_CONFIG_HOME=/u01/oracle/crs

  2012-12-28 11:25:33.822: [RACG][3600347632] [18308][3600347632][ora.bo2dbp.vip]: clsrcexecut:

  cmd = /u01/oracle/crs/bin/racgeut -e _USR_ORA_DEBUG=0 54 /u01/oracle/crs/bin/racgvip start bo2dbp

  2012-12-28 11:25:33.822: [RACG][3600347632] [18308][3600347632][ora.bo2dbp.vip]: clsrcexecut: rc = 1, time = 3.210s

  2012-12-28 11:25:37.063: [RACG][3600347632] [18308][3600347632][ora.bo2dbp.vip]: clsrcexecut:

  env ORACLE_CONFIG_HOME=/u01/oracle/crs

  2012-12-28 11:25:37.063: [RACG][3600347632] [18308][3600347632][ora.bo2dbp.vip]: clsrcexecut:

  cmd = /u01/oracle/crs/bin/racgeut -e _USR_ORA_DEBUG=0 54 /u01/oracle/crs/bin/racgvip check bo2dbp

  2012-12-28 11:25:37.063: [RACG][3600347632] [18308][3600347632][ora.bo2dbp.vip]: clsrcexecut: rc = 1, time = 3.240s

  2012-12-28 11:25:37.063: [RACG][3600347632] [18308][3600347632][ora.bo2dbp.vip]: end for resource =

  ora.bo2dbp.vip, action = start, status = 1, time = 6.490s

  #从上面的日志可知还是那个网络接口名的问题

  #也可以看出action = start, status = 1, time = 6.490s 这个地方应该是Target为Online,而实际上State为offline

  #网络接口层也改了,那就是这个eth3还在OCR中没有被更新,接下来尝试更新

  bo2dbp:/u01/oracle/crs/bin # ./srvctl modify nodeapps -n bo2dbp -A 192.168.7.61/255.255.255.0/bond1

  #对第二个节点上也采用相同的方式来更新

  bo2dbs:~ # /u01/oracle/crs/bin/srvctl modify nodeapps -n bo2dbs -A 192.168.7.62/255.255.255.0/bond1

  #此时成功启动vip

  oracle@bo2dbp:~> crs_start ora.bo2dbp.vip

  Attempting to start `ora.bo2dbp.vip` on member `bo2dbp`

  Start of `ora.bo2dbp.vip` on member `bo2dbp` succeeded.

  #接下来查看ons的日志信息

  oracle@bo2dbp:/u01/oracle/crs/log/bo2dbp/racg> tail -20 ora.bo2dbp.ons.log

  ............

  onscfg[0]

  {node = bo2dbp.2gotrade.com, port = 6200}

  Adding remote host bo2dbp.2gotrade.com:6200

  onscfg[1]

  {node = bo2dbs.2gotrade.com, port = 6200}

  Adding remote host bo2dbs.2gotrade.com:6200

  ons is n

  2012-12-28 11:00:49.345: [ RACG][2554102256] [19071][2554102256][ora.bo2dbp.ons]: ot running ...

  2012-12-28 11:00:49.345: [ RACG][2554102256] [19071][2554102256][ora.bo2dbp.ons]:

  clsrcexecut: env ORACLE_CONFIG_HOME=/u01/oracle/crs

  2012-12-28 11:00:49.345: [ RACG][2554102256] [19071][2554102256][ora.bo2dbp.ons]: clsrcexecut:

  cmd = /u01/oracle/crs/bin/racgeut -e _USR_ORA_DEBUG=0 540 /u01/oracle/crs/bin/onsctl ping

  2012-12-28 11:00:49.345: [ RACG][2554102256] [19071][2554102256][ora.bo2dbp.ons]: clsrcexecut: rc = 1, time = 0.210s

  2012-12-28 11:00:49.346: [ RACG][2554102256] [19071][2554102256][ora.bo2dbp.ons]: end

  for resource = ora.bo2dbp.ons, action = start, status = 1, time = 7.560s

  2012-12-28 11:00:55.661: [ RACG][368746992] [19812][368746992][ora.bo2dbp.ons]: onsctl: shutting down ons daemon ...

  CONNECT: Connection refused

  Number of onsconfiguration retrieved, numcfg = 2

  onscfg[0]

  {node = bo2dbp.2gotrade.com, port = 6200}

  Adding remote host bo2dbp.2gotrade.com:6200

  onscfg[1]

  {node = bo2dbs.2gotrade.com,

  2012-12-28 11:00:55.661: [ RACG][368746992] [19812][368746992][ora.bo2dbp.ons]: port = 6200}

  ...............

  #关于ons的这个错误,记得之前有类似的情形,之前是在安装的时候碰到的,是由于没有本地回环造成的。

  #这个问题再次出现了,由于从原来的配置复制/etc/hosts时不小心那个地方被注释掉了,汗......

  #尝试启动ons成功

  oracle@bo2dbp:~> crs_start ora.bo2dbp.ons

  Attempting to start `ora.bo2dbp.ons` on member `bo2dbp`

  Start of `ora.bo2dbp.ons` on member `bo2dbp` succeeded.

  #Author : Robinson

  5、小结

  a、对于RAC环境下虚拟机的迁移之后,首选需要更改各个虚拟机配置文件中所有磁盘的路径(local disk,asmdisk,ocr,votingdisk)

  b、添加虚拟机方式使用了复制方式(移动方式不确定)后将导致网卡发生变化,主要是保证MAC地址唯一

  c、需要对网络重新进行配置,如果希望使用原来的网络接口名,则重命名或命令行下修改相关配置文件使得保留原有网卡名

  d、如果使用了X window方式修改网络配置,应注意hosts文件是否发生相应的修改

  e、如果使用了新的网络接口名或者新的IP地址,应该重新配置集群网络层

  f、同时也需要将新网络接口名或者新的IP地址更新的OCR

  g、最后一句,从日志来分析与解决问题是源头,是快速定位问题的最佳途径

  (以上内容摘于网络,如有侵权,请告之,将第一时间删除)