Hmm osnet-incorporation.res exists, and is a week old
% ls -l usr/src/pkg/packages.i386/osnet-incorporation.res
-rw-r--r-- 1 jk usr 49425 Jul 31 16:18
usr/src/pkg/packages.i386/osnet-incorporation.res
And pkg.depotd shows "osnet-incorporation"
Index Action Value Package
pkg.fmri set on-nightly/consolidation/osnet/osnet-incorporation
consolidation/osnet/osnet-incorporat...@0.5.11,5.11-0.146:20100804T191120Z
But it seems that it didn't have it's version number updated?
==============================================
Two weeks ago, with b145, the detail information for
osnet-incorporation was:
Name: consolidation/osnet/osnet-incorporation
Summary: OS/Net consolidation incorporation
Publisher: on-nightly
Version: 0.5.11
Build Release: 5.11
Branch: 0.146
Packaging Date: 21. Juli 2010, 17:54:58 Uhr
Size: 0.00 B
FMRI:
pkg:/consolidation/osnet/osnet-incorporat...@0.5.11,5.11-0.146:20100721T175458Z
==============================================
With the b146 bits, it is
Name: consolidation/osnet/osnet-incorporation
Summary: OS/Net consolidation incorporation
Publisher: on-nightly
Version: 0.5.11
Build Release: 5.11
Branch: 0.146
Packaging Date: 4. August 2010, 19:11:20 Uhr
Size: 0.00 B
FMRI:
pkg:/consolidation/osnet/osnet-incorporat...@0.5.11,5.11-0.146:20100804T191120Z
==============================================
The packaging date and the FMRI timestamp have changed,
but the branch "0.146" hasn't changed. Is that a problem?
2010/8/5 Mark J. Nelson <mark.j.nel...@oracle.com>:
> The usual symptom is "no updates available," though what you're seeing
> could be the same bug.
>
> The reason it doesn't show up with incremental builds is because
> usr/src/pkg/packages.$ARCH/osnet-incorporation.res still exists from a
> previous build, and that triggers publication
>
> One way to verify would be to run pkg.depotd on top of repo.redist and
> see if it contains osnet-incorporation. If not, it's this bug.
>
> --Mark
>
> On 08/ 4/10 04:56 PM, Juergen Keil wrote:
>> 2010/8/4 Mark J. Nelson <mark.j.nel...@oracle.com>:
>>
>>> Starting with the integration of the cset on Monday:
>>>
>>>> changeset: 12995:f69f91784e5c
>>>> 6945225 need to remove workarounds for recently changed IPS behavior
>>>
>>> ...and ending with the integration of this cset today:
>>>
>>>> Changeset: 13011:dc5824d1233f
>>>> 6974588 synthetic packages not being published in clobber builds
>>>
>>> the package builds from onnv will produce unusable repositories.
>>>
>>> Exception: if you do incremental package builds, the repositories will
>>> be fine. It is only for clean workspaces or clobber builds that they're
>>> broken.
>>
>> Hmm, what happens exactly when you try to use such an unusable repository
>> with onu ?
>>
>> I'm external to Oracle / Sun.
>>
>> I performed several incremental build attempts (changeset 13001, tagged
>> onnv_146) until I got the correct pkg bits installed, and the correct closed
>> binaries installed. The b146 source did build ok with nightly -i, it seems.
>>
>> But an onu using b134 as the base BE, and the compiled b146 packages,
>> installed almost nothing:
>>
>> # /opt/onbld/bin/onu -O -t osol_146_onu -d /files/snv_146_jk/packages/i386/
>> nightly-nd
>>
>> [04/Aug/2010:22:54:58] ENGINE Listening for SIGHUP.
>> [04/Aug/2010:22:54:58] ENGINE Listening for SIGTERM.
>> [04/Aug/2010:22:54:58] ENGINE Listening for SIGUSR1.
>> [04/Aug/2010:22:54:58] ENGINE Bus STARTING
>> [04/Aug/2010:22:54:58] ENGINE Started monitor thread '_TimeoutMonitor'.
>> [04/Aug/2010:22:54:58] ENGINE Serving on 0.0.0.0:13000
>> [04/Aug/2010:22:54:58] ENGINE Bus STARTED
>> [04/Aug/2010:22:54:58] INDEX Updating search indices
>> PHASE ACTIONS
>> Entfernphase 54/54
>> [04/Aug/2010:22:55:49] INDEX Search indexes updated and available.
>> PHASE ACTIONS
>> Entfernphase 46/46
>> Installationsphase 47/47
>> Aktualisierungsphase 22/22
>> osol_146_onu wurde erfolgreich aktualisiert
>> [04/Aug/2010:22:58:40] ENGINE Caught signal SIGTERM.
>> [04/Aug/2010:22:58:40] ENGINE Bus STOPPING
>> [04/Aug/2010:22:58:40] ENGINE HTTP Server
>> cherrypy._cpwsgi_server.CPWSGIServer(('0.0.0.0', 13000)) shut down
>>
>>
>> In the past a few thousand files were deleted, created and updated by onu.
>>
>> But this time it deletes, creates and updates less than 50 files. And the
>> new
>> BE boots a b134 kernel.
>>
>>
>> Is that the same bug?
>>
>> If that's the same bug: Why did I run into this bug with incremental builds?
>
>
_______________________________________________
on-discuss mailing list
on-discuss@opensolaris.org
http://mail.opensolaris.org/mailman/listinfo/on-discuss
