Hi Jan.
I reviewed your code changes and they look reasonable to me.
- I looked at the order that the new entries for
update_dumpadm_nodename() are called in install_finish, which seems to
be correct when I look at the existing call there for the GUI install.
- I'm OK with just pulling out the extra step of a temporary file used
to create dumpadm.conf (since this is a new system anyway and dumps
anyway would be critical to be able to use the system).
My only nit is ict.py / lines 889- 890. Since you're copying
dumpadm.conf from a running system to another BASEDIR, I would suggest
something like this:
dumpadmfile = '/etc/dumpadm.conf'
dumpadmfile_dest = self.BASEDIR + dumpadmfile
... but this is only a suggestion.
Thanks,
Jack
On 04/27/09 07:35, Jan Damborsky wrote:
> Hi,
>
> could I please Joe and another pair of eyes to review changes for
> following blocker ?
>
> 8427 AI doesn't set dump device to ZVOL on installed system
>
> webrev:
> http://cr.opensolaris.org/~dambi/bug-8427
>
> Thank you very much,
> Jan
>
>
> modules affected:
> -----------------
> * ICT - install-finish & ict.py
>
> testing done - please see attached test procedures
> ---------------------------------------------------
>
> [1] LiveCD - regression test
> ----------------------------
> - used 111a ISO with modified bits
>
> [2] x86 & SPARC AI
> ------------------
> - AI images built from http://ipkg.sfbay/dev - based on 111a
> - non-default hostname set in SC manifest
> - verified that after installation, both DUMPADM_DEVICE
> and DUMPADM_SAVDIR parameters are set to correct values
> in /etc/dumpadm.conf on installed system
>
> _______________________________________________
> caiman-discuss mailing list
> caiman-discuss at opensolaris.org
> http://mail.opensolaris.org/mailman/listinfo/caiman-discuss
