DNS 引起经典RAC故障

DNS 引起经典RAC故障

作者：吴伟龙(PrudentWoo)

一、环境介绍：

这是一套四年前部署的RAC系统，之前运行一直很好，没有出过问题，平时基本处于无人管的状态。

OS:Redhat EnterPrise Linux 5.8x86_x64

DB:Oracle DatabaseEnterPrise  11.2.0.4

GI:Oracle Grid Infrastructure11.2.0.4

二、问题描述：

         昨天临近下班接到现场人员故障请求，描述为数据库无法连接，报ORA-12547:TNS:lost  CONNECT。当时第一反应是网络和监听故障，让现场人员进行tnsping和ping都是正常的。

 

三、问题现象：

         我到达现场后，首先查看了数据库的状态，发现数据库实例是停止运行状态，并且从日志中看不出明显报错；

   数据库日志：

Starting up:
Oracle Database 11g Enterprise Edition Release 11.2.0.4.0 - 64bit Production
With the Partitioning, Real Application Clusters, OLAP, Data Mining
and Real Application Testing options.
ORACLE_HOME = /u01/app/oracle/11.2.0.4/product/db_1
System name:	Linux
Node name:	db01
Release:	3.8.13-44.1.1.el6uek.x86_64
Version:	#2 SMP Wed Sep 10 06:10:25 PDT 2014
Machine:	x86_64
VM name:	VMWare Version: 6
Using parameter settings in server-side pfile /u01/app/oracle/11.2.0.4/product/db_1/dbs/initwoo1.ora
System parameters with non-default values:
  processes                = 600
  sessions                 = 922
  spfile                   = "+DATA/woo/spfilewoo.ora"
  nls_language             = "SIMPLIFIED CHINESE"
  nls_territory            = "CHINA"
  memory_target            = 1584M
  control_files            = "+DATA/woo/controlfile/current.260.930748953"
  control_files            = "+FRA01/woo/controlfile/current.256.930748953"
  db_block_size            = 8192
  compatible               = "11.2.0.4.0"
  cluster_database         = TRUE
  db_create_file_dest      = "+DATA"
  db_recovery_file_dest    = "+FRA01"
  db_recovery_file_dest_size= 4407M
  thread                   = 1
  undo_tablespace          = "UNDOTBS1"
  instance_number          = 1
  remote_login_passwordfile= "EXCLUSIVE"
  db_domain                = ""
  dispatchers              = "(PROTOCOL=TCP) (SERVICE=wooXDB)"
  remote_listener          = "scan.prudentwoo.com:1521"
  audit_file_dest          = "/u01/app/oracle/admin/woo/adump"
  audit_trail              = "DB"
  db_name                  = "woo"
  open_cursors             = 300
  diagnostic_dest          = "/u01/app/oracle"
Cluster communication is configured to use the following interface(s) for this instance
  169.254.51.38
  169.254.243.157
cluster interconnect IPC version:Oracle UDP/IP (generic)
IPC Vendor 1 proto 2
Fri Dec 16 15:24:55 2016
USER (ospid: 4044): terminating the instance due to error 119
Instance terminated by USER, pid = 4044

    数据库状态：

[oracle@db01 ~]$ crsctl status res -t
--------------------------------------------------------------------------------
NAME           TARGET  STATE        SERVER                   STATE_DETAILS       
--------------------------------------------------------------------------------
Local Resources
--------------------------------------------------------------------------------
ora.BAK01.dg
               ONLINE  ONLINE       db01                                         
               ONLINE  ONLINE       db02                                         
ora.DATA.dg
               ONLINE  ONLINE       db01                                         
               ONLINE  ONLINE       db02                                         
ora.FRA01.dg
               ONLINE  ONLINE       db01                                         
               ONLINE  ONLINE       db02                                         
ora.LISTENER.lsnr
               ONLINE  ONLINE       db01                                         
               ONLINE  ONLINE       db02                                         
ora.OCR_VOT.dg
               ONLINE  ONLINE       db01                                         
               ONLINE  ONLINE       db02                                         
ora.asm
               ONLINE  ONLINE       db01                     Started             
               ONLINE  ONLINE       db02                     Started             
ora.gsd
               OFFLINE OFFLINE      db01                                         
               OFFLINE OFFLINE      db02                                         
ora.net1.network
               ONLINE  ONLINE       db01                                         
               ONLINE  ONLINE       db02                                         
ora.ons
               ONLINE  ONLINE       db01                                         
               ONLINE  ONLINE       db02                                         
--------------------------------------------------------------------------------
Cluster Resources
--------------------------------------------------------------------------------
ora.LISTENER_SCAN1.lsnr
      1        ONLINE  ONLINE       db02                                         
ora.LISTENER_SCAN2.lsnr
      1        ONLINE  ONLINE       db01                                         
ora.LISTENER_SCAN3.lsnr
      1        ONLINE  ONLINE       db01                                         
ora.cvu
      1        ONLINE  ONLINE       db01                                         
ora.db01.vip
      1        ONLINE  ONLINE       db01                                         
ora.db02.vip
      1        ONLINE  ONLINE       db02                                         
ora.oc4j
      1        ONLINE  ONLINE       db01                                         
ora.scan1.vip
      1        ONLINE  ONLINE       db02                                         
ora.scan2.vip
      1        ONLINE  ONLINE       db01                                         
ora.scan3.vip
      1        ONLINE  ONLINE       db01                                         
ora.woo.db
      1        ONLINE  OFFLINE                               Instance Shutdown   
      2        ONLINE  OFFLINE                               Instance Shutdown   

[oracle@db01 ~]$ srvctl status database -d woo
Instance woo1 is not running on node db01
Instance woo2 is not running on node db02

四、手工带起数据库：

[oracle@db01 trace]$ srvctl start database -d woo
PRCR-1079 : Failed to start resource ora.woo.db
CRS-5017: The resource action "ora.woo.db start" encountered the following error: 
ORA-00119: invalid specification for system parameter REMOTE_LISTENER
ORA-00132: syntax error or unresolved network name 'scan.prudentwoo.com:1521'
. For details refer to "(:CLSN00107:)" in "/u01/app/11.2.0.4/product/grid/log/db02/agent/crsd/oraagent_oracle/oraagent_oracle.log".

CRS-5017: The resource action "ora.woo.db start" encountered the following error: 
ORA-00119: invalid specification for system parameter REMOTE_LISTENER
ORA-00132: syntax error or unresolved network name 'scan.prudentwoo.com:1521'
. For details refer to "(:CLSN00107:)" in "/u01/app/11.2.0.4/product/grid/log/db01/agent/crsd/oraagent_oracle/oraagent_oracle.log".

CRS-2674: Start of 'ora.woo.db' on 'db02' failed
CRS-2674: Start of 'ora.woo.db' on 'db01' failed
CRS-2632: There are no more servers to try to place resource 'ora.woo.db' on that would satisfy its placement policy

    日志信息：

[oracle@db01 trace]$ tail -0f alert_woo1.log 
Fri Dec 16 15:37:08 2016
Starting ORACLE instance (normal)
LICENSE_MAX_SESSION = 0
LICENSE_SESSIONS_WARNING = 0
Initial number of CPU is 2
Private Interface 'eth1:1' configured from GPnP for use as a private interconnect.
  [name='eth1:1', type=1, ip=169.254.51.38, mac=00-0c-29-7c-44-ca, net=169.254.0.0/17, mask=255.255.128.0, use=haip:cluster_interconnect/62]
Private Interface 'eth2:1' configured from GPnP for use as a private interconnect.
  [name='eth2:1', type=1, ip=169.254.243.157, mac=00-0c-29-7c-44-d4, net=169.254.128.0/17, mask=255.255.128.0, use=haip:cluster_interconnect/62]
Public Interface 'eth0' configured from GPnP for use as a public interface.
  [name='eth0', type=1, ip=192.168.84.11, mac=00-0c-29-7c-44-c0, net=192.168.84.0/24, mask=255.255.255.0, use=public/1]
Public Interface 'eth0:1' configured from GPnP for use as a public interface.
  [name='eth0:1', type=1, ip=192.168.84.22, mac=00-0c-29-7c-44-c0, net=192.168.84.0/24, mask=255.255.255.0, use=public/1]
Public Interface 'eth0:3' configured from GPnP for use as a public interface.
  [name='eth0:3', type=1, ip=192.168.84.20, mac=00-0c-29-7c-44-c0, net=192.168.84.0/24, mask=255.255.255.0, use=public/1]
Public Interface 'eth0:5' configured from GPnP for use as a public interface.
  [name='eth0:5', type=1, ip=192.168.84.13, mac=00-0c-29-7c-44-c0, net=192.168.84.0/24, mask=255.255.255.0, use=public/1]
CELL communication is configured to use 0 interface(s):
CELL IP affinity details:
    NUMA status: non-NUMA system
    cellaffinity.ora status: N/A
CELL communication will use 1 IP group(s):
    Grp 0: 
Picked latch-free SCN scheme 3
Using LOG_ARCHIVE_DEST_1 parameter default value as USE_DB_RECOVERY_FILE_DEST
Autotune of undo retention is turned on. 
LICENSE_MAX_USERS = 0
SYS auditing is disabled
Starting up:
Oracle Database 11g Enterprise Edition Release 11.2.0.4.0 - 64bit Production
With the Partitioning, Real Application Clusters, OLAP, Data Mining
and Real Application Testing options.
ORACLE_HOME = /u01/app/oracle/11.2.0.4/product/db_1
System name:	Linux
Node name:	db01
Release:	3.8.13-44.1.1.el6uek.x86_64
Version:	#2 SMP Wed Sep 10 06:10:25 PDT 2014
Machine:	x86_64
VM name:	VMWare Version: 6
Using parameter settings in server-side pfile /u01/app/oracle/11.2.0.4/product/db_1/dbs/initwoo1.ora
System parameters with non-default values:
  processes                = 600
  sessions                 = 922
  spfile                   = "+DATA/woo/spfilewoo.ora"
  nls_language             = "SIMPLIFIED CHINESE"
  nls_territory            = "CHINA"
  memory_target            = 1584M
  control_files            = "+DATA/woo/controlfile/current.260.930748953"
  control_files            = "+FRA01/woo/controlfile/current.256.930748953"
  db_block_size            = 8192
  compatible               = "11.2.0.4.0"
  cluster_database         = TRUE
  db_create_file_dest      = "+DATA"
  db_recovery_file_dest    = "+FRA01"
  db_recovery_file_dest_size= 4407M
  thread                   = 1
  undo_tablespace          = "UNDOTBS1"
  instance_number          = 1
  remote_login_passwordfile= "EXCLUSIVE"
  db_domain                = ""
  dispatchers              = "(PROTOCOL=TCP) (SERVICE=wooXDB)"
  remote_listener          = "scan.prudentwoo.com:1521"
  audit_file_dest          = "/u01/app/oracle/admin/woo/adump"
  audit_trail              = "DB"
  db_name                  = "woo"
  open_cursors             = 300
  diagnostic_dest          = "/u01/app/oracle"
Cluster communication is configured to use the following interface(s) for this instance
  169.254.51.38
  169.254.243.157
cluster interconnect IPC version:Oracle UDP/IP (generic)
IPC Vendor 1 proto 2
Fri Dec 16 15:37:49 2016
USER (ospid: 6043): terminating the instance due to error 119
Instance terminated by USER, pid = 6043

五、问题分析：

         我从启动数据库来看，发现数据库此时无法正常启动，并随着报ORA-00132，日志报error119。

         根据启动提示可以将问题定位到scan，因scan故障引起数据库无法正常启动。

六、检查scan配置信息：

[oracle@db01 ~]$ srvctl config scan
SCAN name: scan.prudentwoo.com, Network: 1/192.168.84.0/255.255.255.0/eth0
SCAN VIP name: scan1, IP: /scan.prudentwoo.com/192.168.84.21
SCAN VIP name: scan2, IP: /scan.prudentwoo.com/192.168.84.22
SCAN VIP name: scan3, IP: /scan.prudentwoo.com/192.168.84.20


[oracle@db01 ~]$ ping 192.168.84.20 -c 2
PING 192.168.84.20 (192.168.84.20) 56(84) bytes of data.
64 bytes from 192.168.84.20: icmp_seq=1 ttl=64 time=0.032 ms
64 bytes from 192.168.84.20: icmp_seq=2 ttl=64 time=0.039 ms

--- 192.168.84.20 ping statistics ---
2 packets transmitted, 2 received, 0% packet loss, time 1000ms
rtt min/avg/max/mdev = 0.032/0.035/0.039/0.006 ms
[oracle@db01 ~]$ ping 192.168.84.21 -c 2
PING 192.168.84.21 (192.168.84.21) 56(84) bytes of data.
64 bytes from 192.168.84.21: icmp_seq=1 ttl=64 time=0.231 ms
64 bytes from 192.168.84.21: icmp_seq=2 ttl=64 time=0.292 ms

--- 192.168.84.21 ping statistics ---
2 packets transmitted, 2 received, 0% packet loss, time 1001ms
rtt min/avg/max/mdev = 0.231/0.261/0.292/0.034 ms
[oracle@db01 ~]$ ping 192.168.84.22 -c 2
PING 192.168.84.22 (192.168.84.22) 56(84) bytes of data.
64 bytes from 192.168.84.22: icmp_seq=1 ttl=64 time=0.024 ms
64 bytes from 192.168.84.22: icmp_seq=2 ttl=64 time=0.034 ms

--- 192.168.84.22 ping statistics ---
2 packets transmitted, 2 received, 0% packet loss, time 999ms
rtt min/avg/max/mdev = 0.024/0.029/0.034/0.005 ms

[oracle@db01 ~]$ ping scan.prudentwoo.com -c 2
ping: unknown host scan.prudentwoo.com

我们可以看到，现在scan对应的三个地址都是通的，说明SCAN的服务是好的，但是pingscan所对应的域名的时候报无法找到主机，无法解析域名，那么下一步可以定位应该是域名服务出问题了。

七、在两台数据库服务器上检查域名(dns)服务，结果是域名服务器没有在这两台数据服务器上：

#check dns client and server:
[oracle@db01 ~]$ /sbin/chkconfig --list|grep named
[oracle@db01 ~]$ ssh db02 '/sbin/chkconfig --list|grep named'
[oracle@db01 ~]$ 

check dns client:
[oracle@db01 ~]$ cat /etc/resolv.conf 
search prudentwoo.com
nameserver 192.168.84.15

八、根据resolv.conf配置找到真正的域名服务器，发现域名域名服务器hang住：

[oracle@db01 ~]$ ping 192.168.84.15 -c 2
PING 192.168.84.15 (192.168.84.15) 56(84) bytes of data.
From 192.168.84.11 icmp_seq=1 Destination Host Unreachable
From 192.168.84.11 icmp_seq=2 Destination Host Unreachable

--- 192.168.84.15 ping statistics ---
2 packets transmitted, 0 received, +2 errors, 100% packet loss, time 3007ms
pipe 2

九、修复域名服务器，现在可以正常解析：

[oracle@db01 ~]$ ping scan.prudentwoo.com -c 2
PING scan.prudentwoo.com (192.168.84.21) 56(84) bytes of data.
64 bytes from scan.prudentwoo.com (192.168.84.21): icmp_seq=1 ttl=64 time=0.494 ms
64 bytes from scan.prudentwoo.com (192.168.84.21): icmp_seq=2 ttl=64 time=0.289 ms

--- scan.prudentwoo.com ping statistics ---
2 packets transmitted, 2 received, 0% packet loss, time 1001ms
rtt min/avg/max/mdev = 0.289/0.391/0.494/0.104 ms

十、再次启动数据库

[oracle@db01 ~]$ srvctl start database -d woo
[oracle@db01 ~]$ srvctl status database -d woo
Instance woo1 is running on node db01
Instance woo2 is running on node db02

[oracle@db01 ~]$ srvctl config database -d woo
Database unique name: woo
Database name: woo
Oracle home: /u01/app/oracle/11.2.0.4/product/db_1
Oracle user: oracle
Spfile: +DATA/woo/spfilewoo.ora
Domain: 
Start options: open
Stop options: immediate
Database role: PRIMARY
Management policy: AUTOMATIC
Server pools: woo
Database instances: woo1,woo2
Disk Groups: DATA,FRA01
Mount point paths: 
Services: 
Type: RAC
Database is administrator managed

能正常启动，故障修复。

 

       从整个问题的处理思路来看该故障不仅考验解决数据库故障能力，同时安装，基本运行原理都有考察到，当然考验更多的应该还是操作系统和DNS服务的深入理解。

      当然我是很庆幸的，出于职业敏感度，一堆报错中瞬间发现问题根源ORA-00132，而没有从其它报错信息入手。