public inbox for linux-xfs@vger.kernel.org
 help / color / mirror / Atom feed
From: "Artur Makówka" <artur.makowka@gmail.com>
To: xfs@oss.sgi.com
Subject: creating new array
Date: Tue, 04 Mar 2008 17:31:12 +0100	[thread overview]
Message-ID: <47CD7950.9040001@gmail.com> (raw)

i want to create new array, because my last one (raid 0) is gone... now 
it is going to be raid 5.

this will be on xfs fs of course. now the problem is i don't really know 
what chunk size i should use, to get the best performance, but i can 
still mount old array read-only and check what is average file size 
there - just don't know how. (doing ls -lR * , adding sizes and dividing 
by number is probably going to take days to complete)

there was some xfs command that showed it, but i really have no idea 
what it was, xfs_info doesnt give me much information

the currect situation is like this:

meta-data=/dev/md0               isize=256    agcount=32, agsize=5723342 
blks
          =                       sectsz=512   attr=0
data     =                       bsize=4096   blocks=183146912, imaxpct=25
          =                       sunit=2      swidth=6 blks
naming   =version 2              bsize=4096
log      =internal               bsize=4096   blocks=32768, version=2
          =                       sectsz=512   sunit=0 blks, lazy-count=0
realtime =none                   extsz=24576  blocks=0, rtextents=0


some of the files are not on it anymore, as i said, it is damaged RAID 0

the system is going to have millions of small files, (up to 3MB), or 
even hundreds of millions (very popular free hosting)

could you advice me any xfs and mdadm options? i plan to use lazyblocks 
option with latest 2.6.24 kernel. any other ideas?

this is going to be RAID 5, the stability is NOT very important, i will 
have full backups, the most important thing is performance. i can get 
hundreds of hits per second at peak times, i was planning to build it 
from 5 * 500GB disks and it will be just partition for users file, 
system is on another disk. (probably on lvm to give me option to resize 
it in future)

fs performance is critical, as i have many thousands of accounts, mostly 
with just small websites, so it's impossible to put it all in RAM.

if you have any xfs/mdadm creating advice (or mounting options), please 
share with me. i installed xfsprogs 2.9.7, and i will compile 2.6.24 
kernel for creation time ( i can even risk using 2.6.25-rcX if there is 
some important thing changed that i need durning array/filesystem 
creation, just not -mm as XFS version from -mm once corrupted some of my 
users files )

                 reply	other threads:[~2008-03-04 16:31 UTC|newest]

Thread overview: [no followups] expand[flat|nested]  mbox.gz  Atom feed

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=47CD7950.9040001@gmail.com \
    --to=artur.makowka@gmail.com \
    --cc=xfs@oss.sgi.com \
    /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