[Vdo-devel] vdo + lvm cache case vdo failed
Sweet Tea Dorminy
sweettea at redhat.com
Wed Apr 10 12:33:09 UTC 2019
Immediately, as part of vdoFormat invocation.
On Wed, Apr 10, 2019 at 8:05 AM Dennis Keefe <dkeefe at redhat.com> wrote:
>
> Question for vdo-devel.
>
> When a VDO volume is created is the magic number written to the backing device immediately or is it written later on, such as, the first time the volume is shutdown?
>
> On Wed, Apr 10, 2019, 1:04 AM 劼磊周 <loli.miko at gmail.com> wrote:
>>
>> sdb sdc are both on the dell h730p controller
>>
>> But I found before reboot . VDO bits was gone
>> [root ~]# hexdump -C -s 0 -n 8 /dev/sdc
>> 00000000 00 00 00 00 00 00 00 00 |........|
>> 00000008
>> [root at image-store21008 ~]# lsblk
>> NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
>> sdb 8:16 0 1.5T 0 disk
>> ├─image-cache_cmeta 253:6 0 20G 0 lvm
>> │ └─image-data 253:4 0 116T 0 lvm /exports/disk0
>> └─image-cache_cdata 253:5 0 1.4T 0 lvm
>> └─image-data 253:4 0 116T 0 lvm /exports/disk0
>> sdc 8:32 0 58.2T 0 disk
>> └─image_store 253:3 0 116T 0 vdo
>> └─image-data_corig 253:7 0 116T 0 lvm
>> └─image-data 253:4 0 116T 0 lvm /exports/disk0
>> sda 8:0 0 931G 0 disk
>> ├─sda2 8:2 0 930G 0 part
>> │ ├─centos-swap 253:1 0 4G 0 lvm [SWAP]
>> │ ├─centos-home 253:2 0 876G 0 lvm /home
>> │ └─centos-root 253:0 0 50G 0 lvm /
>> └─sda1 8:1 0 1G 0 part /boot
>> [root@ ~]#
>>
>> On Wed, Apr 10, 2019 at 1:01 PM Dennis Keefe <dkeefe at redhat.com> wrote:
>>>
>>> The first 8 bites of the vdo device should show "dmvdo001". It doesnt look like the device has a VDO volume on it.
>>>
>>> It looks like sdc and sdb have both changed after reboot. Are sdb and sdc on the same controller?
>>>
>>> On Tue, Apr 9, 2019, 9:43 PM 劼磊周 <loli.miko at gmail.com> wrote:
>>>>
>>>> and
>>>> lsblk
>>>> NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
>>>> sdb 8:16 0 1.5T 0 disk
>>>> ├─image-cache_cmeta 253:6 0 20G 0 lvm
>>>> │ └─image-data 253:4 0 116T 0 lvm /exports/disk0
>>>> └─image-cache_cdata 253:5 0 1.4T 0 lvm
>>>> └─image-data 253:4 0 116T 0 lvm /exports/disk0
>>>> sdc 8:32 0 58.2T 0 disk
>>>> └─image_store 253:3 0 116T 0 vdo
>>>> └─image-data_corig 253:7 0 116T 0 lvm
>>>> └─image-data 253:4 0 116T 0 lvm /exports/disk0
>>>> sda 8:0 0 931G 0 disk
>>>> ├─sda2 8:2 0 930G 0 part
>>>> │ ├─centos-swap 253:1 0 4G 0 lvm [SWAP]
>>>> │ ├─centos-home 253:2 0 876G 0 lvm /home
>>>> │ └─centos-root 253:0 0 50G 0 lvm /
>>>> └─sda1 8:1 0 1G 0 part /boot
>>>>
>>>> this is lsblk info before reboot.
>>>>
>>>> after reboot
>>>> lsblk
>>>> NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
>>>> sdb 8:16 0 1.5T 0 disk
>>>> sdc 8:32 0 58.2T 0 disk
>>>> sda 8:0 0 931G 0 disk
>>>> ├─sda2 8:2 0 930G 0 part
>>>> │ ├─centos-swap 253:1 0 4G 0 lvm [SWAP]
>>>> │ └─centos-root 253:0 0 926G 0 lvm /
>>>> └─sda1 8:1 0 1G 0 part /boot
>>>>
>>>>
>>>>
>>>> On Wed, Apr 10, 2019 at 10:39 AM 劼磊周 <loli.miko at gmail.com> wrote:
>>>>>
>>>>> hi
>>>>> I am use local disk(8TB * 10 raid6) disk
>>>>>
>>>>> hexdump -C -s 0 -n 8 /dev/sdc
>>>>> 00000000 00 00 00 00 00 00 00 00 |........|
>>>>>
>>>>> blkid -p /dev/sdc return empty
>>>>>
>>>>>
>>>>> On Wed, Apr 10, 2019 at 10:27 AM Dennis Keefe <dkeefe at redhat.com> wrote:
>>>>>>
>>>>>> Thank you for the version information.
>>>>>>
>>>>>> What type of storage are you using (local disk, iscsi, multipath)? Can you run these commands on the device?
>>>>>>
>>>>>> hexdump -C -s 0 -n 8 <device and path>
>>>>>>
>>>>>> blkid -p <device and path>
>>>>>>
>>>>>> -Dennis
>>>>>>
>>>>>> On Tue, Apr 9, 2019, 8:21 PM 劼磊周 <loli.miko at gmail.com> wrote:
>>>>>>>
>>>>>>> hi:
>>>>>>> I am using centos 7.6 with kernel 5.0.4-1.el7.elrepo.x86_64
>>>>>>> and i use a patch file for compile with 5.0 kernel
>>>>>>> https://gist.githubusercontent.com/rhawalsh/5ef42b2dbf4084f4f8baa1761fe0d60a/raw/eb124e3f7a370c956e9375937a071cbe904280d1/0001-removed_variable-length-arrays.patch
>>>>>>>
>>>>>>> and vdo version is
>>>>>>> Version information:
>>>>>>> kvdo version: 6.2.0.293
>>>>>>> vdo is the same as kvdo.
>>>>>>>
>>>>>>>
>>>>>>> On Tue, Apr 9, 2019 at 8:53 PM Dennis Keefe <dkeefe at redhat.com> wrote:
>>>>>>>>
>>>>>>>> Interesting. Would you mind providing the kernel version and package version for vdo and kmod-vdo?
>>>>>>>>
>>>>>>>> On Tue, Apr 9, 2019, 4:18 AM 劼磊周 <loli.miko at gmail.com> wrote:
>>>>>>>>>
>>>>>>>>> hello：
>>>>>>>>>
>>>>>>>>> Today i try to use lvm cache on vdo device
>>>>>>>>>
>>>>>>>>> i use the command
>>>>>>>>>
>>>>>>>>> pvcreate /dev/sdb
>>>>>>>>> pvcreate /dev/mapper/image_store
>>>>>>>>> vgcreate image /dev/sdb
>>>>>>>>> vgextend image /dev/mapper/image_store
>>>>>>>>> lvcreate -l 20G -n meta image /dev/sdb
>>>>>>>>> lvcreate -l 100%FREE -n cache image /dev/sdb
>>>>>>>>> lvconvert --type cache-pool --poolmetadata image/meta image/cache
>>>>>>>>>
>>>>>>>>> lvcreate -l 100%FREE -n data image /dev/mapper/image_store
>>>>>>>>> lvconvert --type cache --cachepool image/cache --cachemode writeback image/data
>>>>>>>>> mkfs.xfs -K /dev/mapper/image-data
>>>>>>>>> mount /dev/mapper/image-data /exports/disk0/
>>>>>>>>>
>>>>>>>>> in my case sdb is ssd device
>>>>>>>>> /dev/mapper/image_store is a vdo device
>>>>>>>>>
>>>>>>>>> Everything look fine after reboot.
>>>>>>>>>
>>>>>>>>> We i reboot server. Vdo device cannot be start.
>>>>>>>>>
>>>>>>>>> vdo: utils.FileLock: DEBUG - attempting to lock /run/lock/vdo/_etc_vdoconf.yml.lock in 20s mode exclusive
>>>>>>>>> vdo: utils.FileLock: DEBUG - attempting to lock /run/lock/vdo-config-singletons mode exclusive
>>>>>>>>> vdo: vdo.vdomgmnt.Configuration: DEBUG - Reading configuration from /etc/vdoconf.yml
>>>>>>>>> vdo: utils.FileLock: DEBUG - released lock /run/lock/vdo-config-singletons
>>>>>>>>> vdo: vdo.vdomgmnt.Service.VDOService: INFO - Starting VDO image_store
>>>>>>>>> Starting VDO image_store
>>>>>>>>> vdo: vdo.vdomgmnt.Service.VDOService: DEBUG - No failure requiring recovery for VDO volume image_store
>>>>>>>>> vdo: utils.Command: DEBUG - dmsetup: command failed, exit status 1
>>>>>>>>> vdo: utils.Command: DEBUG - stdout:
>>>>>>>>> vdo: utils.Command: DEBUG - stderr: Device does not exist.
>>>>>>>>> Command failed.
>>>>>>>>> vdo: utils.Command: DEBUG - grep: command succeeded
>>>>>>>>> vdo: utils.Command: DEBUG - stdout: MemAvailable: 130560288 kB
>>>>>>>>> vdo: utils.Command: DEBUG - stderr:
>>>>>>>>> vdo: utils.Command: DEBUG - modprobe: command succeeded
>>>>>>>>> vdo: utils.Command: DEBUG - stdout:
>>>>>>>>> vdo: utils.Command: DEBUG - stderr:
>>>>>>>>> vdo: utils.Command: DEBUG - vdodumpconfig: command failed, exit status 1
>>>>>>>>> vdo: utils.Command: DEBUG - stdout:
>>>>>>>>> vdo: utils.Command: DEBUG - stderr: vdodumpconfig: allocateVDO failed for '/dev/disk/by-id/scsi-364cd98f05bc87f00242af26c0b2df932' with VDO Status: Bad magic number
>>>>>>>>> vdodumpconfig: Could not load VDO from '/dev/disk/by-id/scsi-364cd98f05bc87f00242af26c0b2df932'
>>>>>>>>> vdo: vdo.vdomgmnt.Service.VDOService: ERROR - Could not set up device mapper for image_store
>>>>>>>>> vdo: vdo.vdomgmnt.Configuration: DEBUG - Configuration is clean, not persisting
>>>>>>>>> vdo: utils.FileLock: DEBUG - released lock /run/lock/vdo/_etc_vdoconf.yml.lock
>>>>>>>>> vdo: vdo: ERROR - vdodumpconfig: allocateVDO failed for '/dev/disk/by-id/scsi-364cd98f05bc87f00242af26c0b2df932' with VDO Status: Bad magic number
>>>>>>>>>
>>>>>>>>> Bad magic number ERROR was be shown.
>>>>>>>>>
>>>>>>>>> Thanks
>>>>>>>>>
>>>>>>>>> --
>>>>>>>>> leon.zhou
>>>>>>>>>
>>>>>>>>> _______________________________________________
>>>>>>>>> Vdo-devel mailing list
>>>>>>>>> Vdo-devel at redhat.com
>>>>>>>>> https://www.redhat.com/mailman/listinfo/vdo-devel
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> --
>>>>>>> 周劼磊
>>>>>>>
>>>>>
>>>>>
>>>>>
>>>>> --
>>>>> 周劼磊
>>>>>
>>>>
>>>>
>>>>
>>>> --
>>>> 周劼磊
>>>>
>>
>>
>>
>> --
>> 周劼磊
>>
>
> _______________________________________________
> Vdo-devel mailing list
> Vdo-devel at redhat.com
> https://www.redhat.com/mailman/listinfo/vdo-devel
More information about the vdo-devel
mailing list