All of lore.kernel.org
 help / color / mirror / Atom feed
* xfs tuning for a 830 GB partition (mkfs.xfs options)
@ 2010-12-08 19:32 Abel Coto
  2010-12-08 22:14 ` Stan Hoeppner
  2010-12-09  1:05 ` Dave Chinner
  0 siblings, 2 replies; 5+ messages in thread
From: Abel Coto @ 2010-12-08 19:32 UTC (permalink / raw)
  To: xfs


[-- Attachment #1.1: Type: text/plain, Size: 2629 bytes --]

I want to create a 830 GB partition, to mount as /home in my Centos 5.4
workstation.

Actually i have /home not mounted ,so i would have to create the new
partition in the lvm,format it and use rsync to copy /home directory to it.

for a 830 GB partition that i will use to save my data in general and also
3d / CG projects and renders once finished them what agcount value should be
used. I have read that mkfs default option creates 1 allocation group each
4G , so i understand that for a 830 GB partition agcount should be 208.

It is this correct ?

so i would use for format the partition

mkfs.xfs -l lazy-count=1,version=2,size=128m -i attr=2 -d agcount=208 -L
VolumeName <dev>

I have read that some people instead use a 8GB allocation size , and agcount
Size partition / 8 GB (in my case 104) , but i don't know if that would be
good , as would result in less allocation groups.

i understand if agcount > 208  (for example 416 ) , the allocation size (2GB
for agcount=416) would be less than 4 GB , to have in total 830 GB.

I have read that too much allocation groups are bad , but i don't now if 416
are  too much for a 830 GB partition (and shoukd use 208 to 300) or not.

I will mount the partition with logbufs=8. (i use a 128m journal and
logbufs=8 to improve a bit deleting performance). I will perhaps use noatime
too ,but i have to see if my backup app, uses atime or not.


I have another 1,5 TB HDD where i want to create a 350 GB xfs partition to
save the 3d projects and render output when working with the 3d software i
use (Maya 2011). My thoughts are to use /home to save both music ,
photography,isos,etc also of finished 3d / cg work and iso dvd / tutorials
and video-tutorials ,and this 350 GB partition use it only for data and
projects that i am still working with,and once finished copy to /home.

for this partition , that will handle mostly big files (500 MB to 2-3 GB at
least) what mkfs.xfs options should i use ?

something like:

mkfs.xfs -l lazy-count=1,version=2,size=128m -i attr=2 -d agcount=90 -L
VolumeName <dev>  (350/4 --> 88)

I prefer to have a separate partition to save projects at least when i am
working with them , although 3d modelling is not too much filesystem bounded
, and then copy it to /home , as i think having a good backup strategy shoud
not be a problem to have my music and data along with 3d work finished in
the same partition (3d is for me a hobby,that i like a lot but a hobby)

I will have to compile a custom kernel from kernel.org because my centos
kernel is 2.6.18,and there is no higher one available, only compiling it,
but this shouldn't be a problem.

[-- Attachment #1.2: Type: text/html, Size: 2811 bytes --]

[-- Attachment #2: Type: text/plain, Size: 121 bytes --]

_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: xfs tuning for a 830 GB partition (mkfs.xfs options)
  2010-12-08 19:32 xfs tuning for a 830 GB partition (mkfs.xfs options) Abel Coto
@ 2010-12-08 22:14 ` Stan Hoeppner
       [not found]   ` <AANLkTi=qkM9c5ukdJj7Hqhj8XMWSOGNENLEG9_Uz_DBT@mail.gmail.com>
  2010-12-09  1:05 ` Dave Chinner
  1 sibling, 1 reply; 5+ messages in thread
From: Stan Hoeppner @ 2010-12-08 22:14 UTC (permalink / raw)
  To: xfs

Abel Coto put forth on 12/8/2010 1:32 PM:
> I want to create a 830 GB partition, to mount as /home in my Centos 5.4
> workstation.
> 
> Actually i have /home not mounted ,so i would have to create the new
> partition in the lvm,format it and use rsync to copy /home directory to it.
> 
> for a 830 GB partition that i will use to save my data in general and also
> 3d / CG projects and renders once finished them what agcount value should be
> used. I have read that mkfs default option creates 1 allocation group each
> 4G , so i understand that for a 830 GB partition agcount should be 208.
> 
> It is this correct ?

No that's not correct.

> so i would use for format the partition
> 
> mkfs.xfs -l lazy-count=1,version=2,size=128m -i attr=2 -d agcount=208 -L
> VolumeName <dev>

Do not do this!  If this filesystem will reside on a single physical
disk, format the partition using the XFS defaults.  agcount, and most of
the other options, exists for optimizing parallel performance on striped
RAID or SSD storage systems that have lots of IOPS performance.  These
options are _not_ for use on single disk drives.

AG count is related to number of spindles and/or IOPs throughput, not
the size of the partition.  The mkfs.xfs default for a single drive
filesystem is 4 AGs.  If you specify a value greater than 4 your
performance will suffer.  If you specify 208 AGs it may likely be little
faster than a floppy disk drive.

-- 
Stan

_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: xfs tuning for a 830 GB partition (mkfs.xfs options)
       [not found]   ` <AANLkTi=qkM9c5ukdJj7Hqhj8XMWSOGNENLEG9_Uz_DBT@mail.gmail.com>
@ 2010-12-09  0:27     ` Stan Hoeppner
  0 siblings, 0 replies; 5+ messages in thread
From: Stan Hoeppner @ 2010-12-09  0:27 UTC (permalink / raw)
  To: xfs

Abel Coto put forth on 12/8/2010 5:29 PM:

> By the moment i will use one hard disk for / (on ext3) and /home , and
> the data partition for saving 3d projects working on in another hard
> disk (1,5 tb one , with the 350 GB partition and one of 1 TB for
> backups of /home mostly)
> 
> so agcount should be in both cases 4, as i haven't got a raid (the 1,5
> tb disk won't be in the LVM neither).
> 
> I have read that in xfsprogs 3.1.2, i think, -l lazy-count=1,version=2
> and -i attr=2 is the default option like agcount=4. So  i should only
> specify the -l size=128m if i want a log size of 128m for improve
> deleting performance and improving system in general.

With a single disk filesystem, you can tweak these things forever, and
you will gain almost no performance benefit over the mkfs.xfs defaults,
though you may decrease performance if you don't know exactly what
you're doing, such as with your previous misunderstanding of agcount.

The one thing you can tweak beyond defaults to get improved performance
is using delayed logging, which will increase metadata write performance
substantially.  For example, deletes of large groups (thousands) of
small files will be "massively" faster using delayed logging.  IIRC this
requires kernel 2.6.36 or later.  Simply add "delaylog" to you
/etc/fstab mount options.

> I think in /home the most important thing is read/write seqential
> performance , and in /data (the 350 GB partition) the same.
> 
> I don't think /data would have to handle too many ramdon reads/ writes
> when Maya saves and read data from it, i think would be mostly
> seqential and then allocated in memory.

Again, with a single disk, there's not much you can tweak to increase
general XFS performance, except for metadata writes using delaylog.  For
sequential file reads/writes and random reads/writes, you're at the
mercy of the drive's spindle speed.

XFS options can't improve upon the poor physics of a single spinning
drive.  If you need improved file performance (not metadata), your
options are to add spindles and stripe the data (RAID card or mdadm,
RAID 0/5/6/10), or use a good quality SSD.

-- 
Stan

_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: xfs tuning for a 830 GB partition (mkfs.xfs options)
  2010-12-08 19:32 xfs tuning for a 830 GB partition (mkfs.xfs options) Abel Coto
  2010-12-08 22:14 ` Stan Hoeppner
@ 2010-12-09  1:05 ` Dave Chinner
  2010-12-09  1:34   ` Dave Chinner
  1 sibling, 1 reply; 5+ messages in thread
From: Dave Chinner @ 2010-12-09  1:05 UTC (permalink / raw)
  To: Abel Coto; +Cc: xfs

On Wed, Dec 08, 2010 at 08:32:28PM +0100, Abel Coto wrote:
> I want to create a 830 GB partition, to mount as /home in my Centos 5.4
> workstation.
> 
> Actually i have /home not mounted ,so i would have to create the new
> partition in the lvm,format it and use rsync to copy /home directory to it.
> 
> for a 830 GB partition that i will use to save my data in general and also
> 3d / CG projects and renders once finished them what agcount value should be
> used. I have read that mkfs default option creates 1 allocation group each
> 4G , so i understand that for a 830 GB partition agcount should be 208.
> 
> It is this correct ?

No, you've read something that is at least 5 years out of date. Just
use the defaults - they are already optimised for best performance
in most circumstances.

> I have read that too much allocation groups are bad , but i don't now if 416
> are  too much for a 830 GB partition (and shoukd use 208 to 300) or not.

Any more than 4 AGs on a single spindle is bad for performance. AGs
can be up to 1TB in size, so you're going to get 4 as the default
for your 830GB partition.

> I will mount the partition with logbufs=8. (i use a 128m journal and
> logbufs=8 to improve a bit deleting performance). I will perhaps use noatime
> too ,but i have to see if my backup app, uses atime or not.

logbufs=8 is the default. Also, the default atime option is relatime
which has pretty much zero compareeed to noatime, so once again just
use the defaults.

> for this partition , that will handle mostly big files (500 MB to 2-3 GB at
> least) what mkfs.xfs options should i use ?

The defaults.

Cheers,

Dave.
-- 
Dave Chinner
david@fromorbit.com

_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: xfs tuning for a 830 GB partition (mkfs.xfs options)
  2010-12-09  1:05 ` Dave Chinner
@ 2010-12-09  1:34   ` Dave Chinner
  0 siblings, 0 replies; 5+ messages in thread
From: Dave Chinner @ 2010-12-09  1:34 UTC (permalink / raw)
  To: Abel Coto; +Cc: xfs

On Thu, Dec 09, 2010 at 12:05:31PM +1100, Dave Chinner wrote:
> On Wed, Dec 08, 2010 at 08:32:28PM +0100, Abel Coto wrote:
> > I want to create a 830 GB partition, to mount as /home in my Centos 5.4
> > workstation.
> > 
> > Actually i have /home not mounted ,so i would have to create the new
> > partition in the lvm,format it and use rsync to copy /home directory to it.
> > 
> > for a 830 GB partition that i will use to save my data in general and also
> > 3d / CG projects and renders once finished them what agcount value should be
> > used. I have read that mkfs default option creates 1 allocation group each
> > 4G , so i understand that for a 830 GB partition agcount should be 208.
> > 
> > It is this correct ?
> 
> No, you've read something that is at least 5 years out of date. Just
> use the defaults - they are already optimised for best performance
> in most circumstances.

Seeing as I say this so often:

http://xfs.org/index.php/XFS_FAQ#Q:_I_want_to_tune_my_XFS_filesystems_for_.3Csomething.3E

Cheers,

Dave.
-- 
Dave Chinner
david@fromorbit.com

_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs

^ permalink raw reply	[flat|nested] 5+ messages in thread

end of thread, other threads:[~2010-12-09  1:37 UTC | newest]

Thread overview: 5+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2010-12-08 19:32 xfs tuning for a 830 GB partition (mkfs.xfs options) Abel Coto
2010-12-08 22:14 ` Stan Hoeppner
     [not found]   ` <AANLkTi=qkM9c5ukdJj7Hqhj8XMWSOGNENLEG9_Uz_DBT@mail.gmail.com>
2010-12-09  0:27     ` Stan Hoeppner
2010-12-09  1:05 ` Dave Chinner
2010-12-09  1:34   ` Dave Chinner

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.