在db2 purescale 环境中，模拟 主CF 故障，备CF接管失败？

在4台物理机上搭建db2 purescale 测试环境
系统环境：
物理主机资源：CPU：16C MEM:32G
操作系统版本：suse 11 sp4
数据库版本：v10.5fp10_linuxx64_universal_fixpack

架构图：

测试前状态

db2inst1@db2sr01:~> db2instance -list
ID TYPE STATE HOME_HOST CURRENT_HOST ALERT PARTITION_NUMBER LOGICAL_PORT NETNAME

0 MEMBER STARTED db2sr02 db2sr02 NO 0 0 db2sr02
1 MEMBER STARTED db2sr04 db2sr04 NO 0 0 db2sr04
128 CF PRIMARY db2sr01 db2sr01 NO - 0 db2sr01
129 CF PEER db2sr03 db2sr03 NO - 0 db2sr03

HOSTNAME STATE INSTANCE_STOPPED ALERT

db2sr03 ACTIVE NO NO
db2sr01 ACTIVE NO NO
db2sr04 ACTIVE NO NO
db2sr02 ACTIVE NO NO

手工执行 db2stop CF 128 #### db2sr03能正常接管CF，MEM1,MEM2能正常读写数据库

db2inst1@db2sr02:~> db2instance -list
ID TYPE STATE HOME_HOST CURRENT_HOST ALERT PARTITION_NUMBER LOGICAL_PORT NETNAME

0 MEMBER STARTED db2sr02 db2sr02 NO 0 0 db2sr02
1 MEMBER STARTED db2sr04 db2sr04 NO 0 0 db2sr04
128 CF PEER db2sr01 db2sr01 NO - 0 db2sr01
129 CF PRIMARY db2sr03 db2sr03 NO - 0 db2sr03

HOSTNAME STATE INSTANCE_STOPPED ALERT

db2sr03 ACTIVE NO NO
db2sr01 ACTIVE NO NO
db2sr04 ACTIVE NO NO
db2sr02 ACTIVE NO NO
db2inst1@db2sr02:~>

测试db2sr01宕机

db2inst1@db2sr01:~> db2instance -list
ID TYPE STATE HOME_HOST CURRENT_HOST ALERT PARTITION_NUMBER LOGICAL_PORT NETNAME

0 MEMBER STARTED db2sr02 db2sr02 NO 0 0 db2sr02
1 MEMBER STARTED db2sr04 db2sr04 NO 0 0 db2sr04
128 CF PRIMARY db2sr01 db2sr01 NO - 0 db2sr01
129 CF PEER db2sr03 db2sr03 NO - 0 db2sr03

HOSTNAME STATE INSTANCE_STOPPED ALERT

db2sr03 ACTIVE NO NO
db2sr01 ACTIVE NO NO
db2sr04 ACTIVE NO NO
db2sr02 ACTIVE NO NO
db2inst1@db2sr01:~>

拔掉 db2sr01电源后，状态如下，MEM1,MEM2无法读写数据库，db2sr03无法接管CF

db2inst1@db2sr03:~> db2instance -list
ID TYPE STATE HOME_HOST CURRENT_HOST ALERT PARTITION_NUMBER LOGICAL_PORT NETNAME

0 MEMBER STARTED db2sr02 db2sr02 NO 0 0 db2sr02
1 MEMBER STARTED db2sr04 db2sr04 NO 0 0 db2sr04
128 CF ERROR db2sr01 db2sr01 YES - 0 db2sr01
129 CF PEER db2sr03 db2sr03 NO - 0 db2sr03

HOSTNAME STATE INSTANCE_STOPPED ALERT

db2sr03 ACTIVE NO NO
db2sr02 ACTIVE NO NO
db2sr04 ACTIVE NO NO
db2sr01 INACTIVE NO YES
There is currently an alert for a member, CF, or host in the data-sharing instance. For more information on the alert, its impact, and how to clear it, run the following command: 'db2cluster -cm -list -alert'.
db2inst1@db2sr03:~>

在未开启db2sr01系统，CF状态一直保持ERROR状态，数据库不可用，对db2sr01进行加电后正常进入系统后，状态如下：

db2inst1@db2sr02:~> db2instance -list
ID TYPE STATE HOME_HOST CURRENT_HOST ALERT PARTITION_NUMBER LOGICAL_PORT NETNAME

0 MEMBER STARTED db2sr02 db2sr02 NO 0 0 db2sr02
1 MEMBER STARTED db2sr04 db2sr04 NO 0 0 db2sr04
128 CF PEER db2sr01 db2sr01 NO - 0 db2sr01
129 CF PRIMARY db2sr03 db2sr03 NO - 0 db2sr03

HOSTNAME STATE INSTANCE_STOPPED ALERT

db2sr03 ACTIVE NO NO
db2sr01 ACTIVE NO NO
db2sr04 ACTIVE NO NO
db2sr02 ACTIVE NO NO

只要4台主机的中的某一台重启，共享的GPFS文件系统都会卡大约10s以上。

日志如下：

自己猜测的原因：

1、会不会是GPFS文件系统引起，切换不成功。

后端的iscsi存储也是通过x86服务器搭建的，通过本地磁盘划分分区，映射给4台数据库服务器，本地磁盘不支持SCSI-3协议，导致GPFS文件系统出现故障时，不能快速切换。

/usr/lpp/mmfs/bin/mmchconfig usePersistentReserve=yes （修改参数，报错磁盘不支持）

有什么办法能解决？需要修改参数？恳请大家帮忙分析一下原因。