Re: Backup using dd

The skip is only necessary when you go to restore your backup onto a raw device.

The best thing you can do is just create a small database on a raw device and experiment with it. Even if you have a single 10M datafile for the SYSTEM tablespace, you can get a feel for how things should work.

-Kevin

"mako" <makoml_at_jcom.hone.ne.jp> wrote in message news:7cNs7.1835$kn.219077_at_news1.rdc1.ky.home.ne.jp...
> The skip is only needed on
>
> - AIX 4K
> - Dec unix 64K
>
> So, we don't need to skip on Linux as far as I know.
>
> Please refer to news threads
> "How to convert RAW <-> File datafiles of tablespace?" (2001/9/20)
>
> And you have better to try backup files with dd is valid or not
> by replacing datafiles from original to backuped files while instance is
> down.
>
> Hope this helps
>
> Mako
>
>
>
> "ctemp" <ctemp01_at_hotmail.com> wrote in message
> news:4d44e6cf.0109270948.23ede0e8_at_posting.google.com...
> > I am trying to backup a database which datafiles are on raw device. It
> > is a 8.1.7 database on Redhat 6.2. The following command supposed to
> > be used:
> >
> > dd if=/dev/... of=/opt/... bs=<block size> skip=<header block>
> >
> > Question: how many header blocks I need to specify here e.g. bs=512,
> > skip=???
> >
> > Any help info will be appreciated.
> >
> > Thanks,
> > Yu
>
Received on Thu Sep 27 2001 - 17:11:44 CDT