Frank Middleton wrote:

> Set PKG_DUMP_STATS, but it didn't work because the install failed due to
> 13533. Neither 13533 nor the notes in the announcement said to add user
> upnp /before/ doing the update!   pkg is  asking for a bug to be filed,
> but presumably this is a known problem, so was the Traceback expected?

The traceback on a missing user is a known bug.

> Oddly, there already was a upnp entry in /etc/groups. May I suggest that
> 13533 be amended to say that "user upnp should be added before doing an
> image-update else the update will fail with a Python traceback"?

Feel free to do just that, if you like.

> Were the two warnings (see below) expected and are they harmless?

The non-empty directory warnings?  They are likely harmless; you should
check the lost+found directories where they were moved and make sure that
nothing there is important.  As far as anyone told the packaging system,
they weren't.

> Is there an ETA for 6353?

No.  It's not targeted for the next release, as aside from a few recent
hiccups, it's something that can be done manually, even if unsupported.
Full support is going to require the rework of /var/pkg that's pending, for
better support for zones, caching client servers, and service discovery.

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

Reply via email to