On Thu, 2003-12-11 at 22:18, Joshua Schmidlkofer wrote:
> On Thu, 2003-12-11 at 12:52, Yanick Quirion wrote:
> > Hi all,
> >
> > I'm using XFS for the first time with kernel 2.4.22 and LVM2. I just want
> > to know what are the best options to use to format an XFS partition. I will
> > have an Oracle Database on this partition and a FTP site, so the size of
> > the file will be big.
> >
> > Thanks for your help.
> >
> > Regards,
> >
>
> Have you checked out the FAQ?
> http://oss.sgi.com/projects/xfs/faq.html
>
> Also, the man page for mkfs.xfs is very usefull.
>
> The only things that I have heard are:
>
> a. On large volumes keep dcount lower than default.
> mkfs.xfs -d agcount=x
xfsprogs-2.6.0 should do that out of the box. I recently created a 1.6TB
fs which has agcount=32, agsize=13340000, while a 1.7 TB volume created
a year ago has agcount=431, agsize=1048576.
> * agcount has to align with sunit and swidth but I don't com
>
> b. On RAID arrangements, if your RAID is hardware use:
> mkfs.xfs -d suinit=<raid block size>k,swidth=<(n) disks>
> * I also saw one post that said on RAID5 use
> * swith=<(n-1) disks>
>
> c. Increase the size of the log:
> mkfs.xfs -l size=32m
> * maybe someone of greater knowledge can comment on
> * determining what size is reasonable.
>
> d. Mount with more logbufs:
> o logbufs=4 or logbufs=8, this increases (from 2) the number of
> in memory log buffers. This means you can have more active
> transactions at once, and can still perform metadata changes
> while the log is being synced to disk. The flip side of this
> is that the amount of metadata changes which may be lost on
> crash is greater.
>
> e. evaluate volume creation options with the '-N' switch. Take note of
> the output to help decide if what's being done looks optimal.
>
>
>
|