10gRAC系列之三-VIP虚拟地址的可用性增强

10gVIP
在节点二进行测试
1.尝试Down掉VIP，发现VIP会自动重新加载
卸载VIP地址,这里172.16.19.212是VIP地址:

# ifconfig -a
lo0: flags=1000849 mtu 8232 index 1
inet 127.0.0.1 netmask ff000000
bge0: flags=1000843 mtu 1500 index 2
inet 172.16.19.12 netmask ffffff00 broadcast 172.16.19.255
ether 0:3:ba:3c:2a:d
bge0:1: flags=1040843 mtu 1500 index 2
inet 172.16.19.212 netmask ffffff00 broadcast 172.16.19.255
bge1: flags=1000843 mtu 1500 index 3
inet 192.168.18.12 netmask ffffff00 broadcast 192.168.18.255
ether 0:3:ba:3c:2a:e
bge2: flags=1000843 mtu 1500 index 4
inet 192.168.19.12 netmask ffffff00 broadcast 192.168.19.255
ether 0:3:ba:3c:2a:f
# ifconfig bge0:1 down

等待几秒钟,发现VIP又被自动挂接:

# ifconfig -a
lo0: flags=1000849 mtu 8232 index 1
inet 127.0.0.1 netmask ff000000
bge0: flags=1000843 mtu 1500 index 2
inet 172.16.19.12 netmask ffffff00 broadcast 172.16.19.255
ether 0:3:ba:3c:2a:d
bge0:1: flags=1040842 mtu 1500 index 2
inet 172.16.19.212 netmask ffffff00 broadcast 172.16.19.255
bge0:2: flags=1040843 mtu 1500 index 2
inet 172.16.19.212 netmask ffffff00 broadcast 172.16.19.255
bge1: flags=1000843 mtu 1500 index 3
inet 192.168.18.12 netmask ffffff00 broadcast 192.168.18.255
ether 0:3:ba:3c:2a:e
bge2: flags=1000843 mtu 1500 index 4
inet 192.168.19.12 netmask ffffff00 broadcast 192.168.19.255
ether 0:3:ba:3c:2a:f

2.停掉节点2，VIP自动切换至节点1
停掉节点2主机:

# reboot
Disconnected; connection lost (Connection closed.).
Connection to 172.16.19.12 closed.

观察节点1的数据库日志，可以发现数据库迅速检测到失效，自动进行了实例恢复:

Fri Dec 29 11:28:25 2006
Trace dumping is performing id=[cdmp_20061229111759]
Fri Dec 29 11:28:57 2006
Reconfiguration started (old inc 2, new inc 4)
List of nodes:
0
Global Resource Directory frozen
* dead instance detected - domain 0 invalid = TRUE
Communication channels reestablished
Master broadcasted resource hash value bitmaps
Non-local Process blocks cleaned out
Fri Dec 29 11:28:57 2006
LMS 0: 1 GCS shadows cancelled, 1 closed
Fri Dec 29 11:28:57 2006
LMS 1: 2 GCS shadows cancelled, 2 closed
Set master node info
Submitted all remote-enqueue requests
Dwn-cvts replayed, VALBLKs dubious
All grantable enqueues granted
Post SMON to start 1st pass IR
Fri Dec 29 11:28:57 2006
LMS 0: 5401 GCS shadows traversed, 0 replayed
Fri Dec 29 11:28:57 2006
LMS 1: 5532 GCS shadows traversed, 0 replayed
Fri Dec 29 11:28:57 2006
Submitted all GCS remote-cache requests
Fri Dec 29 11:28:57 2006
Instance recovery: looking for dead threads
Fri Dec 29 11:28:57 2006
Post SMON to start 1st pass IR
Fix write in gcs resources
Fri Dec 29 11:28:57 2006
Beginning instance recovery of 1 threads
Reconfiguration complete
Fri Dec 29 11:28:58 2006
parallel recovery started with 2 processes
Fri Dec 29 11:28:58 2006
Started redo scan
Fri Dec 29 11:28:59 2006
Completed redo scan
221 redo blocks read, 55 data blocks need recovery
Fri Dec 29 11:29:00 2006
Started redo application at
Thread 2: logseq 3, block 63414
Fri Dec 29 11:29:00 2006
Recovery of Online Redo Log: Thread 2 Group 3 Seq 3 Reading mem 0
Mem# 0 errs 0: /dev/vx/rdsk/xxfdg/redo201
Fri Dec 29 11:29:00 2006
Completed redo application
Fri Dec 29 11:29:00 2006
Completed instance recovery at
Thread 2: logseq 3, block 63635, scn 665056
46 data blocks read, 55 data blocks written, 221 redo blocks read
Switch log for thread 2 to sequence 4

同时节点2的VIP在节点1上加载:
# hostname
db210-rac1.hurray.com.cn

# ifconfig -a
lo0: flags=1000849 mtu 8232 index 1
inet 127.0.0.1 netmask ff000000
bge0: flags=1000843 mtu 1500 index 2
inet 172.16.19.11 netmask ffffff00 broadcast 172.16.19.255
ether 0:3:ba:3b:7b:bf
bge0:1: flags=1040843 mtu 1500 index 2
inet 172.16.19.211 netmask ffffff00 broadcast 172.16.19.255
bge0:2: flags=1040843 mtu 1500 index 2
inet 172.16.19.212 netmask ffffff00 broadcast 172.16.19.255
bge1: flags=1000843 mtu 1500 index 3
inet 192.168.18.11 netmask ffffff00 broadcast 192.168.18.255
ether 0:3:ba:3b:7b:c0
bge2: flags=1000843 mtu 1500 index 4
inet 192.168.19.11 netmask ffffff00 broadcast 192.168.19.255
ether 0:3:ba:3b:7b:c1
bge2:1: flags=1000843 mtu 1500 index 4
inet 172.16.19.18 netmask ffffff00 broadcast 172.16.19.255

3.节点2恢复后，VIP会自动切换回节点2
节点1恢复正常:

bash-2.05$ ifconfig -a
lo0: flags=1000849 mtu 8232 index 1
inet 127.0.0.1 netmask ff000000
bge0: flags=1000843 mtu 1500 index 2
inet 172.16.19.11 netmask ffffff00 broadcast 172.16.19.255
bge0:1: flags=1040843 mtu 1500 index 2
inet 172.16.19.211 netmask ffffff00 broadcast 172.16.19.255
bge1: flags=1000843 mtu 1500 index 3
inet 192.168.18.11 netmask ffffff00 broadcast 192.168.18.255
bge2: flags=1000843 mtu 1500 index 4
inet 192.168.19.11 netmask ffffff00 broadcast 192.168.19.255
bge2:1: flags=1000843 mtu 1500 index 4
inet 172.16.19.18 netmask ffffff00 broadcast 172.16.19.255

此时数据库日志记录如下重配置过程:

Fri Dec 29 11:32:33 2006
Reconfiguration started (old inc 4, new inc 6)
List of nodes:
0 1
Global Resource Directory frozen
Communication channels reestablished
Master broadcasted resource hash value bitmaps
Non-local Process blocks cleaned out
Fri Dec 29 11:32:34 2006
LMS 0: 0 GCS shadows cancelled, 0 closed
Fri Dec 29 11:32:34 2006
LMS 1: 0 GCS shadows cancelled, 0 closed
Set master node info
Submitted all remote-enqueue requests
Dwn-cvts replayed, VALBLKs dubious
All grantable enqueues granted
Fri Dec 29 11:32:40 2006
LMS 0: 5464 GCS shadows traversed, 3035 replayed
Fri Dec 29 11:32:40 2006
LMS 1: 5620 GCS shadows traversed, 3081 replayed
Fri Dec 29 11:32:40 2006
Submitted all GCS remote-cache requests
Post SMON to start 1st pass IR
Fix write in gcs resources
Reconfiguration complete

-The End-