multipath 路径切换
来源：互联网 发布：逆战混沌16连炮数据 编辑：程序博客网 时间：2024/06/05 00:59
首先查看多路径设备
[root@elt1 ~]# [root@elt1 ~]# multipath -llmpathe (3600c0ff00026402cf4cf5d5601000000) dm-1 DotHill,DH3000size=47G features='0' hwhandler='0' wp=rw|-+- policy='round-robin 0' prio=1 status=enabled| `- 3:0:0:0 sdb 8:16 active ready running`-+- policy='round-robin 0' prio=1 status=active `- 4:0:0:0 sdi 8:128 active ready runningmpathd (3600c0ff00026402cf5cf5d5601000000) dm-0 DotHill,DH3000size=47G features='0' hwhandler='0' wp=rw|-+- policy='round-robin 0' prio=1 status=enabled| `- 3:0:0:1 sdc 8:32 active ready running`-+- policy='round-robin 0' prio=1 status=active `- 4:0:0:1 sdj 8:144 active ready runningmpathc (3600c0ff00026402cdbc95d5601000000) dm-4 DotHill,DH3000size=1.8T features='0' hwhandler='0' wp=rw|-+- policy='round-robin 0' prio=1 status=enabled| `- 3:0:0:2 sdd 8:48 active ready running`-+- policy='round-robin 0' prio=1 status=active `- 4:0:0:2 sdk 8:160 active ready runningmpathb (3600c0ff00026402cdbc95d5602000000) dm-3 DotHill,DH3000size=1.8T features='0' hwhandler='0' wp=rw|-+- policy='round-robin 0' prio=1 status=enabled| `- 3:0:0:3 sde 8:64 active ready running`-+- policy='round-robin 0' prio=1 status=active `- 4:0:0:3 sdl 8:176 active ready runningmpathh (3600c0ff00026402ce0c95d5604000000) dm-6 DotHill,DH3000size=1.8T features='0' hwhandler='0' wp=rw|-+- policy='round-robin 0' prio=1 status=active| `- 3:0:0:6 sdh 8:112 active ready running`-+- policy='round-robin 0' prio=1 status=enabled `- 4:0:0:6 sdo 8:224 active ready runningmpathg (3600c0ff00026402ce0c95d5603000000) dm-5 DotHill,DH3000size=1.8T features='0' hwhandler='0' wp=rw|-+- policy='round-robin 0' prio=1 status=enabled| `- 3:0:0:5 sdg 8:96 active ready running`-+- policy='round-robin 0' prio=1 status=active `- 4:0:0:5 sdn 8:208 active ready runningmpathf (3600c0ff00026402cdcc95d5601000000) dm-2 DotHill,DH3000size=1.8T features='0' hwhandler='0' wp=rw|-+- policy='round-robin 0' prio=1 status=enabled| `- 3:0:0:4 sdf 8:80 active ready running`-+- policy='round-robin 0' prio=1 status=active `- 4:0:0:4 sdm 8:192 active ready running[root@elt1 ~]# 再查看多路径设备的配置
multipathd> show configdefaults { verbosity 2 polling_interval 5 udev_dir "/dev" multipath_dir "/lib64/multipath" path_selector "round-robin 0" path_grouping_policy failover getuid_callout "/lib/udev/scsi_id --whitelisted --device=/dev/%n" prio const features "0" path_checker directio failback manual rr_min_io 1000 rr_min_io_rq 1 max_fds 1048576 rr_weight uniform queue_without_daemon no hwtable_regex_match no flush_on_last_del no user_friendly_names no find_multipaths no log_checker_err always retain_attached_hw_handler no detect_prio no reload_readwrite no replace_wwid_whitespace no force_sync no config_dir "/etc/multipath/conf.d" delay_watch_checks no delay_wait_checks no}multipaths { multipath { wwid 3600c0ff00026402cdbc95d5602000000 alias mpathb } multipath { wwid 3600c0ff00026402cdbc95d5601000000 alias mpathc } multipath { wwid 3600c0ff00026402cf5cf5d5601000000 alias mpathd } multipath { wwid 3600c0ff00026402cf4cf5d5601000000 alias mpathe } multipath { wwid 3600c0ff00026402cdcc95d5601000000 alias mpathf } multipath { wwid 3600c0ff00026402ce0c95d5603000000 alias mpathg } multipath { wwid 3600c0ff00026402ce0c95d5604000000 alias mpathh }}multipathd> 在上面的配置中，路径分组策略参数path_grouping_policy的值为failover，这是一种主动/被动的多路径设备冗余策略，使用这种路径分组策略，一个多路径设备下面的每条路径都会单独分为一组（多路径设备下面有多少条路径就有多少个分组），基于优先权的算法，其中一组会被选为活动的路径，选出的分组中的路径可以对存储进行数据的读写，其他分组中的路径则处于非活动状态。只有活动路径出现故障无法访问存储时，才会从其他非活动分组中再选出一个分组作为活动路径。
这里用多路径设备mpathh来演示多路径切换，将一条非活动的路径切换为活动路径，原来的活动路径则会自动变为非活动状态，首先查看多路径设备mpathh的路径及路径的状态：
[root@elt1 ~]# multipathd -kmultipathd> show map mpathh topologympathh (3600c0ff00026402ce0c95d5604000000) dm-6 DotHill,DH3000size=1.8T features='0' hwhandler='0' wp=rw|-+- policy='round-robin 0' prio=1 status=active| `- 3:0:0:6 sdh 8:112 active ready running`-+- policy='round-robin 0' prio=1 status=enabled `- 4:0:0:6 sdo 8:224 active ready runningmultipathd> 根据上面的显示，多路径设备mpathh下面有两条路径，一条为sdh，一条为sdo。sdo的status=enabled表示此路径为非活动状态；sdh的status=active表示此路径为活动状态。当对多路径设备mpathh进行数据读写时，多路径程序将会调用sdh对存储进行读写，下面来做个测试：
1.利用dd命令测试多路径设备的数据读写
[root@elt1 ~]# dd if=/dev/mapper/mpathh of=test bs=1M count=1024010240+0 records in10240+0 records out10737418240 bytes (11 GB) copied, 31.7054 s, 339 MB/s[root@elt1 ~]# 在执行dd命令时打开另一窗口查看sdh和sdo的读写：
[root@elt1 ~]# iostat -m 1 20|grep -E "sdh|sdo|Device"Device: tps MB_read/s MB_wrtn/s MB_read MB_wrtnsdh 0.80 0.08 0.05 30801 21504sdo 2.31 0.18 0.23 71738 93184Device: tps MB_read/s MB_wrtn/s MB_read MB_wrtnsdh 549.00 64.50 0.00 64 0sdo 0.00 0.00 0.00 0 0Device: tps MB_read/s MB_wrtn/s MB_read MB_wrtnsdh 2568.00 307.00 0.00 307 0sdo 0.00 0.00 0.00 0 0Device: tps MB_read/s MB_wrtn/s MB_read MB_wrtnsdh 2559.00 306.38 0.00 306 0sdo 0.00 0.00 0.00 0 0Device: tps MB_read/s MB_wrtn/s MB_read MB_wrtnsdh 2554.00 306.25 0.00 306 0sdo 0.00 0.00 0.00 0 0Device: tps MB_read/s MB_wrtn/s MB_read MB_wrtnsdh 2578.00 312.38 0.00 312 0sdo 0.00 0.00 0.00 0 0Device: tps MB_read/s MB_wrtn/s MB_read MB_wrtnsdh 2575.00 308.25 0.00 308 0sdo 0.00 0.00 0.00 0 0Device: tps MB_read/s MB_wrtn/s MB_read MB_wrtnsdh 2557.00 305.65 0.00 305 0sdo 0.00 0.00 0.00 0 0Device: tps MB_read/s MB_wrtn/s MB_read MB_wrtnsdh 2570.00 309.72 0.00 309 0sdo 0.00 0.00 0.00 0 0Device: tps MB_read/s MB_wrtn/s MB_read MB_wrtnsdh 2557.00 306.00 0.00 306 0sdo 1.00 0.00 0.00 0 0Device: tps MB_read/s MB_wrtn/s MB_read MB_wrtnsdh 2551.00 305.25 0.00 305 0sdo 0.00 0.00 0.00 0 0Device: tps MB_read/s MB_wrtn/s MB_read MB_wrtnsdh 2554.00 305.75 0.00 305 0sdo 0.00 0.00 0.00 0 0Device: tps MB_read/s MB_wrtn/s MB_read MB_wrtnsdh 2581.00 303.00 0.00 303 0sdo 0.00 0.00 0.00 0 0Device: tps MB_read/s MB_wrtn/s MB_read MB_wrtnsdh 2563.00 307.12 0.00 307 0sdo 0.00 0.00 0.00 0 0Device: tps MB_read/s MB_wrtn/s MB_read MB_wrtnsdh 2586.00 305.12 0.00 305 0sdo 0.00 0.00 0.00 0 0Device: tps MB_read/s MB_wrtn/s MB_read MB_wrtnsdh 2584.00 307.75 0.00 307 0sdo 0.00 0.00 0.00 0 0Device: tps MB_read/s MB_wrtn/s MB_read MB_wrtnsdh 2572.00 309.50 0.00 309 0sdo 0.00 0.00 0.00 0 0Device: tps MB_read/s MB_wrtn/s MB_read MB_wrtnsdh 2574.00 311.12 0.00 311 0sdo 0.00 0.00 0.00 0 0Device: tps MB_read/s MB_wrtn/s MB_read MB_wrtnsdh 2564.00 305.38 0.00 305 0sdo 0.00 0.00 0.00 0 0Device: tps MB_read/s MB_wrtn/s MB_read MB_wrtnsdh 2581.00 306.88 0.00 306 0sdo 0.00 0.00 0.00 0 0[root@elt1 ~]# 从上面的iostat命令输出结果可以看出，sdo并没有对存储进行读写，只有shh对存储进行了读操作。下面继续演示将sdo切换为活动路径，然后再对多路径设备mpathh进行数据读写测试。
2.多路径设备路径切换
再次查看多路径设备mpathh的路径状态：
multipathd> show multipath mpathh topologympathh (3600c0ff00026402ce0c95d5604000000) dm-6 DotHill,DH3000size=1.8T features='0' hwhandler='0' wp=rw|-+- policy='round-robin 0' prio=1 status=active| `- 3:0:0:6 sdh 8:112 active ready running`-+- policy='round-robin 0' prio=1 status=enabled `- 4:0:0:6 sdo 8:224 active ready runningmultipathd> 多路径设备mpathh下面有两组路径，第1组路径下面有一条路径sdh，第二组下面有一条路径sdh，当前活动的路径为sdo，也就是第二组。每个组都有一个组编号，第1组的组编号为1，第二组的组编号为2，当路径切换时会用到组编号。下面演示路径切换。
[root@elt1 ~]# multipathd -kmultipathd> helpmultipath-tools v0.4.9 (04/04, 2009)CLI commands reference: list|show paths list|show paths format $format list|show status list|show daemon list|show maps|multipaths list|show maps|multipaths status list|show maps|multipaths stats list|show maps|multipaths format $format list|show maps|multipaths topology list|show topology list|show map|multipath $map topology list|show config list|show blacklist list|show devices list|show wildcards add path $path remove|del path $path add map|multipath $map remove|del map|multipath $map switch|switchgroup map|multipath $map group $group reconfigure suspend map|multipath $map resume map|multipath $map resize map|multipath $map disablequeueing map|multipath $map restorequeueing map|multipath $map disablequeueing maps|multipaths restorequeueing maps|multipaths reinstate path $path fail path $path paths count forcequeueing daemon restorequeueing daemon quit|exit map|multipath $map getprstatus map|multipath $map setprstatus map|multipath $map unsetprstatusmultipathd> 将将第2组路径切换为活动路径：
multipathd> switch map mpathh group 2okmultipathd> 查看路径状态：
mpathh (3600c0ff00026402ce0c95d5604000000) dm-6 DotHill,DH3000size=1.8T features='0' hwhandler='0' wp=rw|-+- policy='round-robin 0' prio=1 status=enabled| `- 3:0:0:6 sdh 8:112 active ready running`-+- policy='round-robin 0' prio=1 status=enabled `- 4:0:0:6 sdo 8:224 active ready running发现第2组路径并没有活动，这可能是有延时的原因，路径状态还没有变成活动状态，下面用dd命令对多路径设备进行读操作并观察数据读写情况。
3.路径切换后的数据读写测试
执行下面的命令，对多路径设备mpathh进行读数据操作：
[root@elt1 ~]# dd if=/dev/mapper/mpathh of=test bs=1M count=1024010240+0 records in10240+0 records out10737418240 bytes (11 GB) copied, 72.2445 s, 149 MB/s[root@elt1 ~]# 执行dd命令的同时打开另一窗口观察数据读写：
[root@elt1 ~]# iostat -m 1 20|grep -E "sdh|sdo|Device"Device: tps MB_read/s MB_wrtn/s MB_read MB_wrtnsdh 0.84 0.08 0.06 30797 21504sdo 2.19 0.16 0.24 61494 93184Device: tps MB_read/s MB_wrtn/s MB_read MB_wrtnsdh 0.00 0.00 0.00 0 0sdo 686.00 70.13 0.00 70 0Device: tps MB_read/s MB_wrtn/s MB_read MB_wrtnsdh 0.00 0.00 0.00 0 0sdo 3128.00 324.50 0.00 324 0Device: tps MB_read/s MB_wrtn/s MB_read MB_wrtnsdh 0.00 0.00 0.00 0 0sdo 3085.00 322.88 0.00 322 0Device: tps MB_read/s MB_wrtn/s MB_read MB_wrtnsdh 0.00 0.00 0.00 0 0sdo 3164.00 327.62 0.00 327 0Device: tps MB_read/s MB_wrtn/s MB_read MB_wrtnsdh 0.00 0.00 0.00 0 0sdo 3012.00 312.88 0.00 312 0Device: tps MB_read/s MB_wrtn/s MB_read MB_wrtnsdh 0.00 0.00 0.00 0 0sdo 2584.00 279.50 0.00 279 0Device: tps MB_read/s MB_wrtn/s MB_read MB_wrtnsdh 0.00 0.00 0.00 0 0sdo 3064.00 314.88 0.00 314 0Device: tps MB_read/s MB_wrtn/s MB_read MB_wrtnsdh 0.00 0.00 0.00 0 0sdo 3041.00 315.62 0.00 315 0Device: tps MB_read/s MB_wrtn/s MB_read MB_wrtnsdh 0.00 0.00 0.00 0 0sdo 2193.00 236.88 0.00 236 0Device: tps MB_read/s MB_wrtn/s MB_read MB_wrtnsdh 0.00 0.00 0.00 0 0sdo 602.00 63.62 0.00 63 0Device: tps MB_read/s MB_wrtn/s MB_read MB_wrtnsdh 0.00 0.00 0.00 0 0sdo 1139.00 122.37 0.00 122 0Device: tps MB_read/s MB_wrtn/s MB_read MB_wrtnsdh 0.00 0.00 0.00 0 0sdo 3175.00 348.25 0.00 348 0Device: tps MB_read/s MB_wrtn/s MB_read MB_wrtnsdh 0.00 0.00 0.00 0 0sdo 3211.00 350.25 0.00 350 0Device: tps MB_read/s MB_wrtn/s MB_read MB_wrtnsdh 0.00 0.00 0.00 0 0sdo 3152.00 336.75 0.00 336 0Device: tps MB_read/s MB_wrtn/s MB_read MB_wrtnsdh 0.00 0.00 0.00 0 0sdo 3138.00 340.50 0.00 340 0Device: tps MB_read/s MB_wrtn/s MB_read MB_wrtnsdh 0.00 0.00 0.00 0 0sdo 1930.00 202.38 0.00 202 0Device: tps MB_read/s MB_wrtn/s MB_read MB_wrtnsdh 0.00 0.00 0.00 0 0sdo 2851.00 303.12 0.00 303 0Device: tps MB_read/s MB_wrtn/s MB_read MB_wrtnsdh 0.00 0.00 0.00 0 0sdo 3147.00 339.00 0.00 339 0Device: tps MB_read/s MB_wrtn/s MB_read MB_wrtnsdh 0.00 0.00 0.00 0 0sdo 2642.00 281.75 0.00 281 0[root@elt1 ~]# 经过多路径切换后的测试，sdo已成为多路径设备mpathh的活动路径，多路径设备mpathh通过sdo对数据进行读写操作。再次查看多路径的状态：
multipathd> show map mpathh topologympathh (3600c0ff00026402ce0c95d5604000000) dm-6 DotHill,DH3000size=1.8T features='0' hwhandler='0' wp=rw|-+- policy='round-robin 0' prio=1 status=enabled| `- 3:0:0:6 sdh 8:112 active ready running`-+- policy='round-robin 0' prio=1 status=active `- 4:0:0:6 sdo 8:224 active ready runningmultipathd> 路径sdo已经处于活动状态，到此多路径切换结束。
在上面的配置中，还有一个非常重要的参数failback，此参数值为manual，当设为manual时，活动路径发生故障后它并不会自动切换到其他路径，需要手动实现路径的切换。如果想在活动路径发生故障后立即切换到其他路径，则failback的值要设置为immediate。failback共有如下几个值：
immediate 值指定立即恢复到包含活跃路径的最高级别路径组群。manual 值指定不需要立即恢复，只有在操作者干预的情况下会发生恢复。followover 值指定当路径组的第一个路径成为活跃路径时应执行自动恢复。这可让节点在另一个节点请求故障修复时不会自动恢复。大于 0 的数字值指定推迟出错切换，以秒表示。默认值为 m anual。在生产环境中如果使用主动/被动的分组策略，最好将failback值设为immediate，这样在发生路径故障后会立即切换到其他正常路径，以保证正常的数据访问。
1 0
- multipath 路径切换
- suse10sp2多路径multipath配置
- 多路径（multipath） 介绍
- redhat multipath多路径 DM
- linux之 multipath 多路径
- Multipath
- multipath
- multipath
- Multipath多路径冗余全解
- multipath多路径配置问题汇总
- Multipath多路径冗余全解析
- [转]Multipath多路径冗余全解析
- Multipath多路径冗余全解
- Multipath多路径冗余全解析
- Linux下多路径multipath配置
- RHEL6.5安装multipath多路径软件
- RHEL6.5安装multipath多路径软件
- multipath多路径实验02-配置多路径软件
- Hibernate与Mybatis对比（转）
- 面试小问题
- java 中的缓存的原理
- ASp.net
- input 子系统开发
- multipath 路径切换
- centos6 伪分布式模式安装hadoop
- C++第七次实验-素数和回文
- openstack m:CentOS 虚拟机动态IP
- 二叉树的层次遍历
- 配置Tomcat8
- 《并行计算》期末总结
- Mockito入门学习(3)–Partial Mock(部分模拟)
- Android错误之--Warning:Gradle version 2.10 is required. Current version is 2.8.
