hpeva4400多路径问题
本帖最后由 ctrl10000unix 于 2012-05-25 10:09 编辑单位一台刀片服务器装了centos6.2，连接hpeva4400存储，存储划分了8个vidisk给服务器，在服务器上用fdisk -l 查看有重复鬼盘，每个重复2次。显示有16个盘。根据网上资料，编辑了multipath.conf，blacklist {
devnode "^sda"
}
defaults {
user_friendly_names yes
path_grouping_policy multibus
failback immediate
no_path_retry fail
}
启动multipath -ll #查看多路径 状态也是正确的
mpathq (3600143800648a20d0000800000310000) dm-7 HP,HSV300
size=500G features='1 queue_if_no_path' hwhandler='0' wp=rw
|-+- policy='round-robin 0' prio=50 status=active
| `- 1:0:1:6 sdo 8:224 active ready running
`-+- policy='round-robin 0' prio=10 status=enabled
`- 1:0:0:6 sdg 8:96active ready running
mpathp (3600143800648a20d0000800000350000) dm-6 HP,HSV300
size=500G features='1 queue_if_no_path' hwhandler='0' wp=rw
|-+- policy='round-robin 0' prio=50 status=active
| `- 1:0:0:7 sdh 8:112 active ready running
`-+- policy='round-robin 0' prio=10 status=enabled
`- 1:0:1:7 sdp 8:240 active ready running
mpatho (3600143800648a20d0000800000290000) dm-5 HP,HSV300
size=500G features='1 queue_if_no_path' hwhandler='0' wp=rw
|-+- policy='round-robin 0' prio=50 status=active
| `- 1:0:1:4 sdm 8:192 active ready running
`-+- policy='round-robin 0' prio=10 status=enabled
`- 1:0:0:4 sde 8:64active ready running
mpathn (3600143800648a20d0000800000390000) dm-4 HP,HSV300
size=500G features='1 queue_if_no_path' hwhandler='0' wp=rw
|-+- policy='round-robin 0' prio=50 status=active
| `- 1:0:1:8 sdq 65:0active ready running
`-+- policy='round-robin 0' prio=10 status=enabled
`- 1:0:0:8 sdi 8:128 active ready running
mpathm (3600143800648a20d0000800000520000) dm-3 HP,HSV300
size=10G features='1 queue_if_no_path' hwhandler='0' wp=rw
|-+- policy='round-robin 0' prio=50 status=active
| `- 1:0:0:1 sdb 8:16active ready running
`-+- policy='round-robin 0' prio=10 status=enabled
`- 1:0:1:1 sdj 8:144 active ready running
mpathl (3600143800648a20d00008000002d0000) dm-2 HP,HSV300
size=500G features='1 queue_if_no_path' hwhandler='0' wp=rw
|-+- policy='round-robin 0' prio=50 status=active
| `- 1:0:0:5 sdf 8:80active ready running
`-+- policy='round-robin 0' prio=10 status=enabled
`- 1:0:1:5 sdn 8:208 active ready running
mpathk (3600143800648a20d0000800000560000) dm-1 HP,HSV300
size=10G features='1 queue_if_no_path' hwhandler='0' wp=rw
|-+- policy='round-robin 0' prio=50 status=active
| `- 1:0:1:2 sdk 8:160 active ready running
`-+- policy='round-robin 0' prio=10 status=enabled
`- 1:0:0:2 sdc 8:32active ready running
mpathj (3600143800648a20d0000800000250000) dm-0 HP,HSV300
size=500G features='1 queue_if_no_path' hwhandler='0' wp=rw
|-+- policy='round-robin 0' prio=50 status=active
| `- 1:0:0:3 sdd 8:48active ready running
`-+- policy='round-robin 0' prio=10 status=enabled
`- 1:0:1:3 sdl 8:176 active ready running
在dev/mapper 下面增加了有mpathjmpathkmpathlmpathmmpathnmpathompathpmpathq这些设备。可以对他们进去fdisk 分区，和格式化，现在问题是，在fdisk -L下面查看居然鬼盘变更多了，原来没启用mulipath 时，看到鬼盘是重复两次，现在启用了多路径服务，磁盘重复了3次，变成3*8=24块盘，多出的8块就是就是上面经过mulitpath聚合的8个盘。因为以前没有接触过多路径这些东西，请教各位大侠，这个现象是正常的吗？我理解应该是启用多路径服务就不应该有鬼盘了呀，怎么用Fdisk 看反而变更多了？？？？ 1.这个称之为鬼盘不合适；
2.现象完全正常，多路径的情况下就是如此；
3.用一个vdisk划给服务器举例，在centos上生成sda和sdb这本来就正常，配置了multipath，再在sda和sdb基础上生产一个mpath*的盘，这个盘就是多路径的盘，你用multipath -ll也看到输出了，这个mpath*的盘包含2个sd*的盘
ps:大多多路径都是这样管理的，也有例外只生成一个整合路径设备的，各家设计不一 deadwind 发表于 2012-05-25 21:06 static/image/common/back.gif
1.这个称之为鬼盘不合适；
2.现象完全正常，多路径的情况下就是如此；
3.用一个vdisk划给服务器举例，在c ...
谢谢，明白了！我以为自己搞得不对，重复了好几次 回复 2# deadwind
学习了
页:
[1]