On Fri, Mar 26, 2010 at 11:36:01AM -0700, Danek Duvall wrote:
> Will Fiveash wrote:
>
> > My next issue is that after reading pkg(1) and beadm(1M) it is still
> > unclear to me whether doing a pkg image-update to update my test system
> > in the scenario described above will result in a new BE with my new debug
> > bits or will just update the existing BE. Can someone clarify this for
> > me?
>
> Currently, pkg image-update *always* creates a new BE if there are any
> changes to be made, while pkg install *never* does, even if the changes
> would require one (in which case it complains).
>
> At some point before the next release, we'd like to have a BE created if
> and only if it's required by the underlying changes, regardless of whether
> image-update or install is doing the work.
The current pkg(1) man page needs to updated to explain exactly when a
BE is created. Currently it states:
With --be-name, rename the newly created boot environment to be the
argument given. This option is only valid if a new boot environment
is created during image update. See also beadm(1m).
and beadm(1m) doesn't mention the term pkg at all.
--
Will Fiveash
Oracle
Austin, TX, USA
Internal Solaris Kerberos/GSS/SASL website: http://kerberos.sfbay.sun.com
http://opensolaris.org/os/project/kerberos/
Note my new work e-mail address: will.fiveash at oracle.com
