Re: hardware raid and xfs

[Alex Mestiashvili <mailatgoogl@gmail.com>]

On 07/13/2012 05:21 PM, Alexander Mestiashvili wrote:
> Hi Debian folks,
>
> Could anyone please explain me why does this warning appear?
>
> mkfs.xfs: Specified data stripe width 640 is not the same as the volume
> stripe width 2048
>
> How does xfs know about volume stripe size width ?
>
> here is the command:
> mkfs.xfs -fb size=4k -d su=64k,sw=5 /dev/mapper/mpath2
> mkfs.xfs: Specified data stripe width 640 is not the same as the volume
> stripe width 2048
> meta-data=/dev/mapper/mpath2     isize=256    agcount=32,
> agsize=12702944 blks
>          =                       sectsz=512   attr=2, projid32bit=0
> data     =                       bsize=4096   blocks=406494128, imaxpct=5
>          =                       sunit=16     swidth=80 blks
> naming   =version 2              bsize=4096   ascii-ci=0
> log      =internal log           bsize=4096   blocks=198496, version=2
>          =                       sectsz=512   sunit=16 blks, lazy-count=1
> realtime =none                   extsz=4096   blocks=0, rtextents=0
>
>
> storage is hardware raid (hp msa g3)
> The LUN allocated to the server is combined from 7 disks in raid6 (I
> know it is not optimal for r6, 6+2 would be better),
> chunk size is 64k
> device is visible as /dev/mapper/mpath2 because of device mapper and
> multipathing.
>
> so the formula for mkfs.xfs seem to be correct: su=chunk size(64k),
> sw=number of data disks (7-2)
>
> Am I missing something ?
>
> Thank you in advance,
> Alex
>
>
>   
Forgot to say:

uname -r : 3.2.0-0.bpo.2-amd64
xfsprogs:  3.1.4

Regards,
Alex