[Linux-cluster] mounting a replicated lun

emmanuel segura emi2fast at gmail.com
Thu Nov 21 17:09:16 UTC 2013


Sorry i missing something, you need to change the physical volume in your
metadata backup

with something like this laHRdG-3NPX-MGMj-TQtV-FkV5-
Yp3M-gSJaj7


2013/11/21 emmanuel segura <emi2fast at gmail.com>

> vgcfgbackup -f /tmp/mylvm.bck c
> change the volume group id: esample
> id = "JGWiR6-J5mh-VnQr-168J-uzj6-RjcD-FtmKQo"
> pvcreate --restorefile /tmp/mylvm.bck --uuid
> "laHRdG-3NPX-MGMj-TQtV-FkV5-Yp3M-gSJaj7" /dev/mapper/mapth5
> vgrename JGWiR6-J5mh-VnQr-168J-uzj6-RjcD-FtmKQo new_vg
>
>
> 2013/11/21 Paras pradhan <pradhanparas at gmail.com>
>
>> It a LVM2 lun.
>>
>> I have presened both luns to a node. Below mpath2 refers to original lun
>> and mapth5 is a replicated lun. both luns are identical. What I need to do
>> is remove mpath2 and use only mpath5. I am still confused what are the
>> steps i need to do to achieve this. I can't do pvcreate  or mkfs.gfs2 I
>> know this.
>>
>> --
>> [root at cvtst3 ~]# multipath -ll
>> mpath2 (36000d310002f1c00000000000000001b) dm-4 COMPELNT,Compellent Vol
>> [size=100G][features=1 queue_if_no_path][hwhandler=0][rw]
>> \_ round-robin 0 [prio=1][active]
>>  \_ 1:0:4:1 sde 8:64  [active][ready]
>>  \_ 1:0:5:1 sdf 8:80  [active][ready]
>>  \_ 4:0:4:1 sdi 8:128 [active][ready]
>>  \_ 4:0:5:1 sdj 8:144 [active][ready]
>> mpath5 (36000d310002fb000000000000000002b) dm-3 COMPELNT,Compellent Vol
>> [size=100G][features=1 queue_if_no_path][hwhandler=0][rw]
>> \_ round-robin 0 [prio=1][active]
>>  \_ 1:0:2:1 sdc 8:32  [active][ready]
>>  \_ 1:0:3:1 sdd 8:48  [active][ready]
>>  \_ 4:0:2:1 sdg 8:96  [active][ready]
>>  \_ 4:0:3:1 sdh 8:112 [active][ready]
>> --
>>
>> Here is the output of pvs if needed. It says duplicate pvs . I am
>> assuming it sees the lvm label(?)
>>
>> --
>> [root at cvtst3 ~]# pvs
>>   Found duplicate PV 0zbIvXwSZbfjD8Pg2KdqvgcHka30P2Ve: using
>> /dev/mpath/36000d310002f1c00000000000000001bp2 not
>> /dev/mpath/36000d310002fb000000000000000002bp2
>>   PV                                             VG         Fmt  Attr
>> PSize  PFree
>>   /dev/mpath/36000d310002f1c00000000000000001bp2 guest_vg_c lvm2 a--
>>  99.97G 996.00M
>>   /dev/sda2                                      VolGroup00 lvm2 a--
>>  67.84G      0
>> ----
>>
>>
>> Thanks
>> Paras.
>>
>>
>> On Thu, Nov 21, 2013 at 4:24 AM, Bryn M. Reeves <bmr at redhat.com> wrote:
>>
>>> On 11/20/2013 10:29 PM, Paras pradhan wrote:
>>> > Since the luns are same in all aspects, will it automatically configure
>>> > itself the new one and mount it to /test?
>>>
>>> As far as mounting goes (this is the same for any file system) it will
>>> depend on how the LUN was configured and used in the first place.
>>>
>>> If you're using something like LVM2 which identifies a device by a label
>>> written to the disk itself then the system will find it and activate
>>> whatever volume groups it contains regardless of the device name or
>>> identifier (WWID).
>>>
>>> If you're using plain multipath or bare devices on the other hand then
>>> the device WWID will be different for the original LUN vs. the
>>> replicated one. This will cause SCSI or multipath device names to change
>>> and if those are used in fstab, scripts or cluster configuration files
>>> then the device will not mount without manual changes to update that
>>> configuration.
>>>
>>> I'm not expert in gfs2 and cluster specifics so you'll need to consider
>>> Bob's advice as well to ensure you get a correct set up for the new
>>> device (and have it consistent across the cluster).
>>>
>>> Regards,
>>> Bryn.
>>>
>>>
>>
>> --
>> Linux-cluster mailing list
>> Linux-cluster at redhat.com
>> https://www.redhat.com/mailman/listinfo/linux-cluster
>>
>
>
>
> --
> esta es mi vida e me la vivo hasta que dios quiera
>



-- 
esta es mi vida e me la vivo hasta que dios quiera
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listman.redhat.com/archives/linux-cluster/attachments/20131121/d00b3715/attachment.htm>


More information about the Linux-cluster mailing list