Tuesday, August 13, 2013

Duplicate physical volumes and multipath missing


Problem:
We ran into a glitch when were trying to extend few database file systems.

Physical volumes shows duplicates and 2 physical volumes multipath missing

[root@mdc2pr002 ~]# cat /etc/redhat-release
Red Hat Enterprise Linux Server release 5.6 (Tikanga)


[root@mdc2pr002 host0]# df -h
Filesystem            Size  Used Avail Use% Mounted on
/dev/mapper/VolGroup00-LogVol00
                      3.9G  872M  2.9G  24% /
/dev/mapper/VolGroup00-LogVol03
                      6.8G  149M  6.3G   3% /tmp
/dev/mapper/VolGroup00-LogVol02
                      9.7G  1.9G  7.4G  20% /var
/dev/mapper/VolGroup00-LogVol01
                      9.9G  2.7G  6.8G  29% /opt
/dev/mapper/VolGroup00-LogVol05
                      9.7G  3.0G  6.3G  32% /home
/dev/mapper/VolGroup00-LogVol04
                      9.7G  2.2G  7.1G  24% /usr
/dev/mapper/VolGroup00-oracleinstall
                       40G   18G   20G  49% /oracle


[root@mdc2pr002 lvm]# pvs
  Found duplicate PV duB4K3d0lBnFTdgAwwfoDx6MxZx6Mdxu: using /dev/sdag1 not /dev/sdq1
  Found duplicate PV FV0I6TmjEEn5kg0hyVF5zEISNTokcdK1: using /dev/sdr1 not /dev/sdb1
  Found duplicate PV jN3eWS2UZ22dmYE37QyV0vSDkuevnTRf: using /dev/sds1 not /dev/sdc1
  Found duplicate PV FV0I6TmjEEn5kg0hyVF5zEISNTokcdK1: using /dev/sdj1 not /dev/sdr1
  Found duplicate PV FV0I6TmjEEn5kg0hyVF5zEISNTokcdK1: using /dev/sdz1 not /dev/sdj1
  Found duplicate PV jN3eWS2UZ22dmYE37QyV0vSDkuevnTRf: using /dev/sdk1 not /dev/sds1
  Found duplicate PV jN3eWS2UZ22dmYE37QyV0vSDkuevnTRf: using /dev/sdaa1 not /dev/sdk1
  PV                             VG         Fmt  Attr PSize   PFree
  /dev/mpath/350002ac0000e0ae2p1            lvm2 a-   333.00G 333.00G
  /dev/mpath/350002ac0000f0ae2p1            lvm2 a-   333.00G 333.00G
  /dev/mpath/350002ac000100ae2p1            lvm2 a-   333.00G 333.00G
  /dev/mpath/350002ac000140ae2p1 VolGroup02 lvm2 a-   499.99G  81.22G    <-- 5="" exported="" mdc2pr002="" nbsp="" p="" to="" x="">  /dev/mpath/350002ac000150ae2p1 VolGroup02 lvm2 a-   499.99G  81.22G    <-- 5="" exported="" mdc2pr002="" nbsp="" p="" to="" x="">  /dev/mpath/350002ac000160ae2p1 VolGroup02 lvm2 a-   499.99G  81.22G    <-- 5="" exported="" mdc2pr002="" nbsp="" p="" to="" x="">  /dev/sda2                      VolGroup00 lvm2 a-   136.00G  45.00G
  /dev/sdaa1                   VolGroup01 lvm2 a-   499.99G 351.16G      <-- 5="" exported="" mdc2pr002="" nbsp="" p="" to="" x="">  /dev/sdz1                      VolGroup01 lvm2 a-   499.99G 252.16G      <-- 5="" exported="" mdc2pr002="" nbsp="" p="" to="" x="">

Multpath shows only 3 x 500GB luns

[root@mdc2pr002 ~]# multipath -ll
mpath2 (350002ac000150ae2) dm-11 3PARdata,VV
[size=500G][features=1 queue_if_no_path][hwhandler=0][rw]    <-- 500gb="" p="">\_ round-robin 0 [prio=0][active]
 \_ 5:0:1:3  sdac 65:192 [active][ready]
 \_ 3:0:0:3  sde  8:64   [active][ready]
 \_ 3:0:1:3  sdm  8:192  [active][ready]
 \_ 5:0:0:3  sdu  65:64  [active][ready]
mpath1 (350002ac000140ae2) dm-10 3PARdata,VV
[size=500G][features=1 queue_if_no_path][hwhandler=0][rw]    <-- 500gb="" p="">\_ round-robin 0 [prio=0][active]
 \_ 5:0:1:2  sdab 65:176 [active][ready]
 \_ 3:0:0:2  sdd  8:48   [active][ready]
 \_ 3:0:1:2  sdl  8:176  [active][ready]
 \_ 5:0:0:2  sdt  65:48  [active][ready]
mpath6 (350002ac0000e0ae2) dm-15 3PARdata,VV
[size=333G][features=1 queue_if_no_path][hwhandler=0][rw]
\_ round-robin 0 [prio=0][active]
 \_ 5:0:1:7  sdag 66:0   [active][ready]
 \_ 3:0:0:7  sdi  8:128  [active][ready]
 \_ 3:0:1:7  sdq  65:0   [active][ready]
 \_ 5:0:0:7  sdy  65:128 [active][ready]
mpath5 (350002ac0000f0ae2) dm-14 3PARdata,VV
[size=333G][features=1 queue_if_no_path][hwhandler=0][rw]
\_ round-robin 0 [prio=0][active]
 \_ 5:0:1:6  sdaf 65:240 [active][ready]
 \_ 3:0:0:6  sdh  8:112  [active][ready]
 \_ 3:0:1:6  sdp  8:240  [active][ready]
 \_ 5:0:0:6  sdx  65:112 [active][ready]
mpath4 (350002ac000100ae2) dm-13 3PARdata,VV
[size=333G][features=1 queue_if_no_path][hwhandler=0][rw]
\_ round-robin 0 [prio=0][active]
 \_ 5:0:1:5  sdae 65:224 [active][ready]
 \_ 3:0:0:5  sdg  8:96   [active][ready]
 \_ 3:0:1:5  sdo  8:224  [active][ready]
 \_ 5:0:0:5  sdw  65:96  [active][ready]
mpath3 (350002ac000160ae2) dm-12 3PARdata,VV
[size=500G][features=1 queue_if_no_path][hwhandler=0][rw]    <-- 500gb="" p="">\_ round-robin 0 [prio=0][active]
 \_ 5:0:1:4  sdad 65:208 [active][ready]
 \_ 3:0:0:4  sdf  8:80   [active][ready]
 \_ 3:0:1:4  sdn  8:208  [active][ready]
 \_ 5:0:0:4  sdv  65:80  [active][ready]

We are not seeing 2 luns (/dev/sdz1, /dev/sdaa1) device via multipath

root@mdc2pr002 ~]# fdisk -l /dev/sdz
Disk /dev/sdz: 536.8 GB, 536870912000 bytes
255 heads, 63 sectors/track, 65270 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes

   Device Boot      Start         End      Blocks   Id  System
/dev/sdz1               1       65270   524281243+  8e  Linux LVM"






Solution:
We do see that we have potential problem with devices because we choose to create physical volumes with raw device instead of multipath devices.
/dev/sdaa1                   VolGroup01 lvm2 a-   499.99G 351.16G      <-- 5="" exported="" mdc2pr002="" nbsp="" p="" to="" x="">/dev/sdz1                      VolGroup01 lvm2 a-   499.99G 252.16G      <-- 5="" exported="" mdc2pr002="" nbsp="" p="" to="" x="">
#0. Step
 We changed the lvm filter /etc/lvm/lvm.conf as
filter = [ "a|/dev/mapper/mpath.*|", "a|/dev/sda$|", "a|/dev/sda1$|", "a|/dev/sda2$|", "r|.*|" ]
instead of default to scan just the multipath devices and /sda and removes all other block devices
#filter = [ "a/.*/" ]


#0. Step
We also changed /etc/multipath.conf as
 user_friendly_names yes


#1. Step
Rebuilding the initrd (RHEL 3, 4, 5)

https://access.redhat.com/site/solutions/1958
[root@mdc2pr002 boot]# uname -r
2.6.18-238.el5

[root@mdc2pr002 boot]# mv /boot/initrd-2.6.18-238.el5.img /boot/initrd-2.6.18-238.el5.img_07302013
[root@mdc2pr002 boot]# mkinitrd -f -v /boot/initrd-2.6.18-238.el5.img 2.6.18-238.el5 $(uname -r)

[root@mdc2pr002 boot]# ls -lat /boot/
total 30051
-rw-------  1 root root 4113444 Jul 30 19:25 initrd-2.6.18-238.el5.img
drwxr-xr-x  4 root root    1024 Jul 30 19:25 .
drwxr-xr-x 34 root root    4096 Jul 30 17:21 ..
drwxr-xr-x  2 root root    1024 Mar 31  2012 grub
-rw-------  1 root root 4113141 Mar 31  2012 initrd-2.6.18-238.el5.img_07302013
-rw-------  1 root root 4113141 Mar 31  2012 initrd-2.6.18-238.el5.img.bak"



#reboot

Now we have multipath friendly name enabled.

[root@mdc2pr002 ~]# pvscan
  PV /dev/mapper/mpath7p1   VG VolGroup01      lvm2 [499.99 GB / 252.16 GB free]
  PV /dev/mapper/mpath0p1   VG VolGroup01      lvm2 [499.99 GB / 351.16 GB free]
  PV /dev/mapper/mpath1p1   VG VolGroup02      lvm2 [499.99 GB / 14.56 GB free]
  PV /dev/mapper/mpath2p1   VG VolGroup02      lvm2 [499.99 GB / 14.56 GB free]
  PV /dev/mapper/mpath3p1   VG VolGroup02      lvm2 [499.99 GB / 14.56 GB free]
  PV /dev/sda2              VG VolGroup00      lvm2 [136.00 GB / 45.00 GB free]
  PV /dev/mapper/mpath5p1                      lvm2 [333.00 GB]
  PV /dev/mapper/mpath4p1                      lvm2 [333.00 GB]
  PV /dev/mapper/mpath6p1                      lvm2 [333.00 GB]"


But we are still missing volume group VolGroup01 and VolGroup02

[root@mdc2pr002 ~]# ls /dev/mapper/
control   mpath1    mpath2p1  mpath4    mpath5p1  mpath7    VolGroup00-LogVol00  VolGroup00-LogVol03  VolGroup00-oracleinstall
mpath0    mpath1p1  mpath3    mpath4p1  mpath6    mpath7p1  VolGroup00-LogVol01  VolGroup00-LogVol04
mpath0p1  mpath2    mpath3p1  mpath5    mpath6p1  mpath8    VolGroup00-LogVol02  VolGroup00-LogVol05

[root@mdc2pr002 host0]# mount -a
mount: special device /dev/VolGroup02/db4-stripe-default does not exist
mount: special device /dev/VolGroup02/db5-stripe-default does not exist
mount: special device /dev/VolGroup02/db6-stripe-default does not exist
mount: special device /dev/VolGroup02/db7-stripe-default does not exist
mount: special device /dev/VolGroup02/db8-stripe-default does not exist



[root@mdc2pr002 ~]# lvscan
  inactive          '/dev/VolGroup01/swap' [99.00 GB] inherit
  inactive          '/dev/VolGroup01/vmdb-stripe-default' [50.00 GB] inherit
  inactive          '/dev/VolGroup01/dbwork-stripe-default' [247.66 GB] inherit
  inactive          '/dev/VolGroup02/db4-stripe-default' [239.06 GB] inherit
  inactive          '/dev/VolGroup02/db5-stripe-default' [269.09 GB] inherit
  inactive          '/dev/VolGroup02/db6-stripe-default' [220.01 GB] inherit
  inactive          '/dev/VolGroup02/db7-stripe-default' [259.07 GB] inherit
  inactive          '/dev/VolGroup02/db8-stripe-default' [469.08 GB] inherit
  ACTIVE            '/dev/VolGroup00/LogVol00' [4.00 GB] inherit
  ACTIVE            '/dev/VolGroup00/LogVol03' [7.00 GB] inherit
  ACTIVE            '/dev/VolGroup00/LogVol02' [10.00 GB] inherit
  ACTIVE            '/dev/VolGroup00/LogVol01' [10.22 GB] inherit
  ACTIVE            '/dev/VolGroup00/LogVol05' [10.00 GB] inherit
  ACTIVE            '/dev/VolGroup00/LogVol04' [10.00 GB] inherit
  ACTIVE            '/dev/VolGroup00/oracleinstall' [39.78 GB] inherit"



We realized these volume groups were inactive state

#2. Step
#vgchange -a y VolGroup01
#vgchange -a y VolGroup02


[root@mdc2pr002 ~]# vgscan
  Reading all physical volumes.  This may take a while...
  Found volume group "VolGroup01" using metadata type lvm2
  Found volume group "VolGroup02" using metadata type lvm2
  Found volume group "VolGroup00" using metadata type lvm2


#3. Step
resize and mount

No comments: