On Thu, Sep 09, 2010 at 01:46:34PM -0700, Darren Reed wrote:
> Should RFEs for pkg go into bugzilla or bugster?
> 
> I think I can see two separate RFEs here:
> 
> - being able to specify the source of the information used to
>   verify the package manifests (and installed data) rather than
>   rely on /var/pkg. That could be either in another BE, a DVD
>   or over the network.
> 
> - being able to specify the root directory under which the
>   package manifests are found that need to be verified.
>   That allows you to boot from a DVD/BE and do "pkg verify"
>   from your hashes against what's in the other directory.
> 
> Together that would potentially let you boot from a 149 DVD
> and verify a 147 install against the server used to install the
> packages from.
> 
> Any others RFEs in this?

I don't know about additonal RFEs, but I think it would help if you
could refine the use case you had in mind.  If you want to boot from a
CD/DVD an verify the contents of a compromised system, that's actually a
different problem than using a selectable set of trust anchors to verify
manifests in an alternate BE.  There's an additional set of metadata
that is used to determine what's installed; it's not just the presence
of a manifest.  If you want to actually verify a compromised system,
we'll need to save a snapshot of this metadata on the read-only media.

IMO, the case for verifying compromised systems duplicates functionality
already implemented in bart(1) and tripwire, but I'm willing to listen
to reasonable arguments to the contrary.

-j
_______________________________________________
pkg-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/pkg-discuss

Reply via email to