Re: [linux-lvm] performance writing to an lv with no file system

["Heinz J . Mauelshagen" <mauelshagen@sistina.com>]

Howard,

the raw throughput you report is likely caused by io bus saturation
and as Patrick mentioned the variation by caching.

Check your io configuration if saturation is an issue.

Regards,
Heinz    -- The LVM Guy --

On Mon, Jun 30, 2003 at 10:21:24PM -0400, Howard Clifford wrote:
> Heinz,
> 
> Thank you for the response.
> 
> My logical volume is striped - 7 ways.
> 
> I am writing to the lv with a simple c program I wrote.
> 
> Each write is 7*stripesize, which in this case is 28 blocks.
> 
> Any thoughts on how I can get top speeds without mounting and unmounting 
>   the volume?
> 
> Thanks,
> Howie
> 
> 
>  > Howard,
> 
>  > what's the LV layout (linear/striped ?).
>  > How do you write to the LV (dd or soemthing).
>  > What's the write pattern (sequential ? block size ?)
> 
>  > 80MB/sec raw sounds like you use a striped LV.
> 
> 
> On Sun, Jun 29, 2003 at 12:08:34PM -0400, Howard Clifford wrote:
>  > I built a 7 disk vg and put 1 lv on it.  I do not want to mount a file
>  > system.
>  > I just want to write directly to the disk device, /dev/vg1/vol1.
>  >
>  > My write performance has an odd behavior.
>  > If after booting I write to the volume I get roughly 175MB/sec.
>  > If after booting I do a mke2fs on the volume, mount it, umount it, then
>  > write to the volume I get roughly 230 MB/sec.
>  >
>  > Any ideas?  How can I get top speed without first mounting then
>  > umounting the volume?
>  > By the way, I tried binding a raw device to the lv and got very bad
>  > performance; around 80 MB/sec.  Would that be expected?
>  >
>  > (I'm using redhat 9.0 -> lvm-1.0.3-12)
>  >
>  > Thanks...
>  >
>  >
>  > _______________________________________________
>  > linux-lvm mailing list
>  > linux-lvm@sistina.com
>  > http://lists.sistina.com/mailman/listinfo/linux-lvm
>  > read the LVM HOW-TO at http://tldp.org/HOWTO/LVM-HOWTO/
> 
> -- 
> 
> Regards,
> Heinz    -- The LVM Guy --
> 
> *** Software bugs are stupid.
>      Nevertheless it needs not so stupid people to solve them ***
> 
> =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
> 
> Heinz Mauelshagen                                 Sistina Software Inc.
> Senior Consultant/Developer                       Am Sonnenhang 11
>                                                    56242 Marienrachdorf
>                                                    Germany
> Mauelshagen@Sistina.com                           +49 2626 141200
>                                                         FAX 924446
> =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
> 
> 
> _______________________________________________
> linux-lvm mailing list
> linux-lvm@sistina.com
> http://lists.sistina.com/mailman/listinfo/linux-lvm
> read the LVM HOW-TO at http://tldp.org/HOWTO/LVM-HOWTO/

*** Software bugs are stupid.
    Nevertheless it needs not so stupid people to solve them ***

=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

Heinz Mauelshagen                                 Sistina Software Inc.
Senior Consultant/Developer                       Am Sonnenhang 11
                                                  56242 Marienrachdorf
                                                  Germany
Mauelshagen@Sistina.com                           +49 2626 141200
                                                       FAX 924446
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-