RAC & ASM -Multipath

[satish]

Dear Erman,

How can i identify the multi path disk device corresponding to my disks?

I ran this command but didnt show any path

[root@erpuatdb1 ~]# multipath -ll
mpathe (360002ac0000000000000003f0001c9fc) dm-45 3PARdata,VV
size=512M features='1 queue_if_no_path' hwhandler='1 alua' wp=rw
`-+- policy='round-robin 0' prio=50 status=active
  |- 2:0:0:4  sdf               8:80    active ready  running
  |- 1:0:0:4  sden              128:240 active ready  running
  |- 2:0:1:4  sdbw              68:160  active ready  running
  `- 1:0:1:4  sdhe              133:64  active ready  running
UAT-REDOLOGS1-01 (360002ac0000000000000005e0001c9fc) dm-19 3PARdata,VV
size=1.0G features='1 queue_if_no_path' hwhandler='1 alua' wp=rw
`-+- policy='round-robin 0' prio=50 status=active
  |- 2:0:0:14 sdp               8:240   active ready  running
  |- 1:0:0:14 sdex              129:144 active ready  running
  |- 2:0:1:14 sdcg              69:64   active ready  running
  `- 1:0:1:14 sdho              133:224 active ready  running
UAT-REDO2LOGS1_DISK2 (360002ac000000000000000c20001c9fc) dm-63 3PARdata,VV
size=1.0G features='1 queue_if_no_path' hwhandler='1 alua' wp=rw
`-+- policy='round-robin 0' prio=50 status=active
  |- 2:0:0:54 sdbd              67:112  active ready  running
  |- 1:0:0:54 sdgl              132:16  active ready  running
  |- 2:0:1:54 sddu              71:192  active ready  running
  `- 1:0:1:54 sdjc              8:352   active ready  running
UAT-DATA13 (360002ac0000000000000009a0001c9fc) dm-49 3PARdata,VV
size=100G features='1 queue_if_no_path' hwhandler='1 alua' wp=rw
`-+- policy='round-robin 0' prio=50 status=active
  |- 2:0:0:41 sdaq              66:160  active ready  running
  |- 1:0:0:41 sdfy              131:64  active ready  running
  |- 2:0:1:41 sddh              70:240  active ready  running
  `- 1:0:1:41 sdip              135:144 active ready  running
mpathd (360002ac0000000000000003e0001c9fc) dm-37 3PARdata,VV
size=512M features='1 queue_if_no_path' hwhandler='1 alua' wp=rw
`-+- policy='round-robin 0' prio=50 status=active
  |- 2:0:0:3  sde               8:64    active ready  running
  |- 1:0:0:3  sdem              128:224 active ready  running
  |- 2:0:1:3  sdbv              68:144  active ready  running
  `- 1:0:1:3  sdhd              133:48  active ready  running
UAT-REDO2LOGS2_DISK5 (360002ac000000000000000ca0001c9fc) dm-7 3PARdata,VV
size=1.0G features='1 queue_if_no_path' hwhandler='1 alua' wp=rw
`-+- policy='round-robin 0' prio=50 status=active
  |- 2:0:0:62 sdbl              67:240  active ready  running
  |- 1:0:0:62 sdgt              132:144 active ready  running
  |- 2:0:1:62 sdec              128:64  active ready  running
  `- 1:0:1:62 sdjk              8:480   active ready  running
UAT-DATA12 (360002ac000000000000000990001c9fc) dm-48 3PARdata,VV
size=100G features='1 queue_if_no_path' hwhandler='1 alua' wp=rw
`-+- policy='round-robin 0' prio=50 status=active
  |- 2:0:0:40 sdap              66:144  active ready  running
  |- 1:0:0:40 sdfx              131:48  active ready  running
  |- 2:0:1:40 sddg              70:224  active ready  running
  `- 1:0:1:40 sdio              135:128 active ready  running

Thank you

[ErmanArslansOracleBlog]

You can make use of oracleasm querydisk and then using major minor disk numbers and device names under /dev, you can map oracleasm devices to physical disk devices.

Check out this blog post->https://blogs.oracle.com/alejandrovargas/mapping-asm-disks-to-physical-devices

[satish]

Thanks for the update erman..

output

[root@erpuatdb1 mapper]# /etc/rc.d/init.d/oracleasm querydisk -d DATA01
Disk "DATA01" is a valid ASM disk on device [253,89]
[root@erpuatdb1 mapper]# ls -l /dev | grep 253, | grep 89
brw-rw----.  1 root disk     253,  89 Apr 23 10:22 dm-89
[root@erpuatdb1 mapper]#


[root@erpuatdb1 mapper]# cd /dev/
[root@erpuatdb1 dev]# ls -lrt dm-89
brw-rw----. 1 root disk 253, 89 Apr 23 10:22 dm-89
[root@erpuatdb1 dev]#


[root@erpuatdb1 disks]# cd /dev/mapper
[root@erpuatdb1 mapper]# ls -lrt

lrwxrwxrwx. 1 root root       8 Apr 23 10:22 UAT-REDOLOGS1-01p1 -> ../dm-82
lrwxrwxrwx. 1 root root       8 Apr 23 10:22 UAT-REDOLOGS1-03p1 -> ../dm-83
lrwxrwxrwx. 1 root root       8 Apr 23 10:22 UAT-REDOLOGS1-04p1 -> ../dm-84
lrwxrwxrwx. 1 root root       8 Apr 23 10:22 UAT-REDOLOGS1-02p1 -> ../dm-85
lrwxrwxrwx. 1 root root       8 Apr 23 10:22 UAT-REDOLOGS2-05p1 -> ../dm-86
lrwxrwxrwx. 1 root root       8 Apr 23 10:22 UAT-REDOLOGS2-02p1 -> ../dm-87
lrwxrwxrwx. 1 root root       8 Apr 23 10:22 UAT-REDOLOGS2-03p1 -> ../dm-88
lrwxrwxrwx. 1 root root       8 Apr 23 10:22 UAT-DATA01p1 -> ../dm-89------------->found here

how can i know if we are using multipath,not able to interpret this output exactly.can you please shed some light what is this dm

[ErmanArslansOracleBlog]

You need to find the WWID s of those disks, and then you need  to check your multipath -ll output again..

some of those disks are already displayed there with their aliases.. But some them are not..

so you need to find the WWIDs and check the multipath outputs accordingly.

here in the example multipath output ->

multipath {
      wwid 35000cca2604a90ac   --> this is the wwid of a disk
      alias HDD_E0_S10_1615499436
}

Example: Finding wwid :

scsi_id -g -u -s /block/sda