Re: Mixed raw and Ufs partitions

Agreed. The Veritas people claim that using their Journaling File system in fast I/O mode works as well as raw I/O. Some of those tablespaces can be combined on the same disk anyway. The DBA guide talks about different scenarios depending on how the DB is used. You typically want as many spindles on different controllers for ultimate performance anyway. 2 disks won't do for this exercise!!

In article <sxpc6.73$7i7.7212_at_nnrp1.sbc.net>,   "Spencer" <spencerp_at_swbell.net> wrote:
> with only two drives, your goal should be to distribute the i/o as
 evenly
> as possible between them. i'm thinking that you're not going to see
> much of an increase in performance by moving to raw partitions, but
> you are certainly going to lose a lot of flexibility. i wouldn't
 recommend
> using raw partitions unless you've got a logical volume manager that
> will allow you to create raw partitions that are appropriately sized.
>
> "Dirk Kiehne" <dirk_at_liberate.com> wrote in message
> news:3A721625.7CECEAC1_at_liberate.com...
> > A question for the experts,
> >
> > I have a development environment:
> >
> > Solaris 8
> > Oracle 8.1.7
> > Our servers interacting with Oracle
> > Two internal SCSI HDs (4.2G and 9.0G)
> >
> > I'm considering switching my Boot Drive (4.2G)
> > to the (9.0G) and installing Oracle on the (4.2G)
> > with Oracle running in raw partions.
> >
> > From my experience installing OPS into Veritas
> > logical volumes on another machine, I would need
> > 13 partitions:
> >
> > system
> > user
> > temp
> > rbs
> > indx
> > tools
> > drsys
> > control1
> > control2
> > redo1_1
> > redo2_2
> > redo2_1
> > redo2_2
> >
> > but I have only potentially 11 available:
> >
> > 4 on the "new" (9.0G): after using /, swap, backup and /liberate
> > 7 on the (4.2G): after using backup
> >
> > Could someone recommend which HD I should put them on and
> > which could reside on Ufs without affecting performance??
> >
> > Thanks in advance,
> > -Dirk
> >
>
>

Sent via Deja.com
http://www.deja.com/ Received on Fri Jan 26 2001 - 21:50:59 CST