From: Zoltan Szecsei <zoltans@geograph.co.za>
To: linux-raid@vger.kernel.org
Subject: Re: mdadm raid6->LVM->ext4 setup
Date: Wed, 17 Nov 2010 09:49:27 +0200 [thread overview]
Message-ID: <4CE38907.8040709@geograph.co.za> (raw)
In-Reply-To: <4CE31641.50909@ziu.info>
On 2010-11-17 01:39, Michal Soltys wrote:
> On 10-11-16 23:17, Luca Berra wrote:
>> On Tue, Nov 16, 2010 at 08:08:44PM +0200, Zoltan Szecsei wrote:
>>>
>>> *** Commands used:
>>> mdadm --create /dev/md0 --metadata=1.2 --auto=md --assume-clean
>>> --bitmap=internal --bitmap-chunk=131072 --chunk=512 --level=6
>>> --raid-devices=8 /dev/sd[abcdefgh]
>>> mdadm --detail --scan >> /etc/mdadm.conf
>>> mdadm --assemble /dev/md0
>> as i said this is not useful
Accepted, but I also understood "not harmfull". I tried it out of
curiosity, it took a nano-second to complete, so didn't try to "undo it".
>>> pvcreate --metadatasize 512 /dev/md0
>> you should use --dataalignment, not --metadatasize, also you should
I checked out man pvcreate & lvm.conf had all the stuff below in it, so
I assumed it would sort itself out.
md_component_detection = 1
md_chunk_alignment = 1
data_alignment_detection = 1
data_alignment_offset_detection = 1
>
> Also with reference to metadatasize - keep in mind the value is
> rounded up to the /next/ 64k boundary (or it was a while ago, perhaps
> it's changed now). So in your case, 512 would really result with 576.
>
> With reference to --dataalignment - there's also --dataalignoffset
> option (although not really important in your setup).
>
> You might consider adjusting
> /sys/block/<yourmddevice>/md/stripe_cache_size to higher values when
> you have everything set. It tends to give quite a performance boost.
I had trouble getting my head around all the calculations, mainly
because I do not know if my sector size of 512 is falsely reported or
not (so I just accepted it as 512).
I am loading the array as we speak, with 200MB to 400MB tiff files. In
the past 11 hours approx 995GB was loaded, so that calculates to a write
speed of 25MB/s
I have no problem stopping & deleting this if you feel that this
represents problems in my setup, but then please can someone do the
alignment calculations & command syntax for me.
Regards to all,
Zoltan
--
===========================================
Zoltan Szecsei PrGISc [PGP0031]
Geograph (Pty) Ltd.
P.O. Box 7, Muizenberg 7950, South Africa.
65 Main Road, Muizenberg 7945
Western Cape, South Africa.
34° 6'16.35"S 18°28'5.62"E
Tel: +27-21-7884897 Mobile: +27-83-6004028
Fax: +27-86-6115323 www.geograph.co.za
===========================================
-----
No virus found in this message.
Checked by AVG - www.avg.com
Version: 10.0.1153 / Virus Database: 424/3261 - Release Date: 11/16/10
--
To unsubscribe from this list: send the line "unsubscribe linux-raid" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
prev parent reply other threads:[~2010-11-17 7:49 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-11-16 18:08 mdadm raid6->LVM->ext4 setup Zoltan Szecsei
2010-11-16 18:16 ` Roman Mamedov
2010-11-16 18:23 ` Zoltan Szecsei
2010-11-16 22:17 ` Luca Berra
2010-11-16 23:39 ` Michal Soltys
2010-11-17 7:49 ` Zoltan Szecsei [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=4CE38907.8040709@geograph.co.za \
--to=zoltans@geograph.co.za \
--cc=linux-raid@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).