[linux-lvm] pvmove failed

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Hello,
 
we've a server (web, ftp, cgi applications) with this hardware:
 
Motherborad SuperMIcro P4DP6
 
Onboard ATI Technologies Inc Rage XL (rev 27) (prog-if 00 [VGA])
 ATI Technologies Inc Rage XL
dual P4 Xeon 2.2 Ghz
4 Gbyte RAM
 
one controller: American Megatrends Inc. MegaRAID (rev 02)
American Megatrends Inc. MegaRAID 475 Express 500 RAID Controller bios ver. 5.2
one controller: SCSI storage controller: Adaptec AHA-3960D / AIC-7899A U160/m (rev 01)
 Adaptec AHA-3960D U160/m
onboard controller:
SCSI storage controller: Adaptec AIC-7899P U160/m (rev 01)
Intel Ethernet Pro 100
 
We've LVM utility software vers. 1.0.5 installed on 2.4.18-10 Readhat Kernel (ver. 7.3, official kernel patched with LVM 1.0.5).
 
The Volume group hosting is so configured:
 
--- Volume group ---
VG Name               hosting
VG Access             read/write
VG Status             available/resizable
VG #                  2
MAX LV                255
Cur LV                3
Open LV               3
MAX LV Size           255.99 GB
Max PV                255
Cur PV                3
Act PV                3
VG Size               34.24 GB
PE Size               4 MB
Total PE              17510
Alloc PE / Size       8744 / 34.16 GB
Free  PE / Size       0 / 0 GB
VG UUID               zsW0nX-g4jP-z1uw-dAEk-TZDP-dxVC-j1Llec
--- Logical volume ---
LV Name                /dev/home/lv
VG Name                home
LV Write Access        read/write
LV Status              available
LV #                   1
# open                 1
LV Size                12 GB
Current LE             3072
Allocated LE           3072
Stripes                2
Stripe size (KByte)    4
Allocation             next free
Read ahead sectors     120
Block device           58:5
 
--- Logical volume ---
LV Name                /dev/hosting/lv
VG Name                hosting
LV Write Access        read/write
LV Status              available
LV #                   1
# open                 1
LV Size                7.16 GB
Current LE             1832
Allocated LE           1832
Stripes                2
Stripe size (KByte)    4
Allocation             next free
Read ahead sectors     120
Block device           58:8
 
--- Logical volume ---
LV Name                /dev/hosting/lv0
VG Name                hosting
LV Write Access        read/write
LV Status              available
LV #                   2
# open                 1
LV Size                7 GB
Current LE             1792
Allocated LE           1792
Stripes                2
Stripe size (KByte)    4
Allocation             next free
Read ahead sectors     120
Block device           58:9
--- Logical volume ---
LV Name                /dev/hosting/lv1
VG Name                hosting
LV Write Access        read/write
LV Status              available
LV #                   3
# open                 1
LV Size                20 GB
Current LE             5120
Allocated LE           5120
Stripes                2
Stripe size (KByte)    4
Allocation             next free
Read ahead sectors     120
Block device           58:10
 
--- Physical volumes ---
PV Name (#)           /dev/sdv1 (1)
PV Status             available / allocatable
Total PE / Free PE    4372 / 0
 
PV Name (#)           /dev/sdw1 (2)
PV Status             available / allocatable
Total PE / Free PE    4372 / 0
The device /dev/sdw1 is an logical array (raid 1) on ami megaraid controller: this device is  in degraded  mode, so we have add two scsi disk with raid software (/dev/md12, one device on onboard controller
and the second on the external AIC 7899); i wolud move the striped device /dev/sdw1 to
new /dev/md12.
    I've add this device (/dev/md12) to volume group hosting with "vgextend" utility.
Note that /dev/sdv1 and /dev/sdw1 are logical striping of volume group hosting
 
In single mode, i've umounted all the logical volume and i've make:
 
/sbin/pvmove -iv /dev/sdw1 /dev/md12
 
The process skipped some PE (I/o errors) but after moved 91 PE from /dev/sdw1 the pvmove command
freezed; after a shutdown of a server i see on the vg hosting that /dev/md12 has 91 PE allocated and i've
mounted on the file system apparently about no errors
    I see also that the PE is on lv logical volume, so i make:
 
/sbin/pvmove -iv -n lv1 /dev/sdw1 /dev/md12
 
(I've tested several type of pvmove comand, but i can't specify a ranges of PE with a logical volume striped).
 
About moved 90 PE the pvmove freezed; reboot the server and /dev/md/12 ha allocated another  90 PE.
 
So go on, and after:
 
/sbin/pvmove -iv -n lv0 /dev/sdw1 /dev/md12 i've moved 89  PE and freezed again:
is this another bug of pvmove utility?
 
Now, the vgdisplay -v hosting is this:
 
 
--- Volume group ---
VG Name               hosting
VG Access             read/write
VG Status             available/resizable
VG #                  2
MAX LV                255
Cur LV                3
Open LV               3
MAX LV Size           255.99 GB
Max PV                255
Cur PV                3
Act PV                3
VG Size               34.24 GB
PE Size               4 MB
Total PE              17510
Alloc PE / Size       8744 / 34.16 GB
Free  PE / Size       8766 / 34.24 GB
VG UUID               zsW0nX-g4jP-z1uw-dAEk-TZDP-dxVC-j1Llec
--- Logical volume ---
LV Name                /dev/home/lv
VG Name                home
LV Write Access        read/write
LV Status              available
LV #                   1
# open                 1
LV Size                12 GB
Current LE             3072
Allocated LE           3072
Stripes                2
Stripe size (KByte)    4
Allocation             next free
Read ahead sectors     120
Block device           58:5
 
--- Logical volume ---
LV Name                /dev/hosting/lv
VG Name                hosting
LV Write Access        read/write
LV Status              available
LV #                   1
# open                 1
LV Size                7.16 GB
Current LE             1832
Allocated LE           1832
Stripes                2
Stripe size (KByte)    4
Allocation             next free
Read ahead sectors     120
Block device           58:8
 
--- Logical volume ---
LV Name                /dev/hosting/lv0
VG Name                hosting
LV Write Access        read/write
LV Status              available
LV #                   2
# open                 1
LV Size                7 GB
Current LE             1792
Allocated LE           1792
Stripes                2
Stripe size (KByte)    4
Allocation             next free
Read ahead sectors     120
Block device           58:9
--- Logical volume ---
LV Name                /dev/hosting/lv1
VG Name                hosting
LV Write Access        read/write
LV Status              available
LV #                   3
# open                 1
LV Size                20 GB
Current LE             5120
Allocated LE           5120
Stripes                2
Stripe size (KByte)    4
Allocation             next free
Read ahead sectors     120
Block device           58:10
 

--- Physical volumes ---
PV Name (#)           /dev/md12 (3)
PV Status             available / allocatable
Total PE / Free PE    8766 / 8499
 
PV Name (#)           /dev/sdv1 (1)
PV Status             available / allocatable
Total PE / Free PE    4372 / 0
 
PV Name (#)           /dev/sdw1 (2)
PV Status             available / allocatable
Total PE / Free PE    4372 / 267
/sbin/pvdisplay -v /dev/md12
 
In this mail is attached the pvdispay output of /dev/md12.

Now the server is up and apparently the data is integrity but i
would safely move all the logical volume
on /dev/md12 or into another device; how can i do this ?
Thanks.

Best Regards.
Davide Di Luzio
System Administrator
-----------------
ALICOM Group     Tel: +39-085-4406901     Fax: +39-085-4406927
 
 
 
 

[Index of Archives]     [Gluster Users]     [Kernel Development]     [Linux Clusters]     [Device Mapper]     [Security]     [Bugtraq]     [MIPS Linux]     [ARM Linux]     [Linux Security]     [Linux RAID]

  Powered by Linux