]>
git.decadent.org.uk Git - dak.git/log
Allan Lyons [Fri, 1 May 2009 16:58:54 +0000 (10:58 -0600)]
Fixed missing DBUpdateError imports.
Signed-off-by: Allan Lyons <Allan_Lyons@wycliffe.ca>
Joerg Jaspert [Thu, 30 Apr 2009 20:38:01 +0000 (22:38 +0200)]
docstring
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Thu, 30 Apr 2009 20:36:55 +0000 (22:36 +0200)]
dinstall
call unchecked with the -p option, it has to run within the locked area.
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Wed, 29 Apr 2009 22:47:47 +0000 (00:47 +0200)]
better handling of that report file
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Wed, 29 Apr 2009 22:41:11 +0000 (00:41 +0200)]
p-u
in case we did get called with a directory, do not complain about non-existing
.changes files. silently exit instead. helpful for calls from within cron, if you
dont want to have an additional otherwise useless if around it
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Wed, 29 Apr 2009 22:39:08 +0000 (00:39 +0200)]
unchecked/dinstall
unchecked stuff now also in functions.
call a few functions from within dinstall, right before we run accepted.
Otherwise, with the new way of NEW, stuff from (o-)p-u-new would take
two dinstall runs before it is from there into the archive proper...
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Wed, 29 Apr 2009 20:40:59 +0000 (22:40 +0200)]
use unchecked.lock, not daily.lock
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Tue, 28 Apr 2009 21:39:50 +0000 (23:39 +0200)]
examine_package
also extract and show README.source, so it exists.
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Tue, 28 Apr 2009 20:02:05 +0000 (22:02 +0200)]
suite_id
if suite doesnt exist dont try to convert a None to int, python doesnt
like this.
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Sun, 26 Apr 2009 11:15:37 +0000 (13:15 +0200)]
NEW
modified process-new and all related stuff so NEW can now also be processed
during (most of) dinstall. The lock where NEW is forbidden is then down to
some 15 til 20 minutes, instead of 2 hours.
For that process-new moves files into a "staging" directory, from which
cron.unchecked will move them into the accepted queue. As c.u only runs when
it is safe to do so this is fine.
The small part where we still are forbidden to do NEW is between the point where
we run process-accepted and right before we start generating Packages files. As
that is exactly the area daily.lock exists for, process-new now locks on that and
simply refuses to do work then.
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Sun, 26 Apr 2009 08:59:04 +0000 (10:59 +0200)]
cron.unchecked
whitespace foo
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Sat, 25 Apr 2009 22:39:22 +0000 (00:39 +0200)]
Merge commit 'buxy/bugfixes' into merge
* commit 'buxy/bugfixes':
dak update-db: fixes to make it work on a fresh install
dak init-db: properly initialize suite_architectures
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Sat, 25 Apr 2009 13:23:08 +0000 (15:23 +0200)]
Merge branch 'merge'
* merge:
HAHA
Joerg Jaspert [Sat, 25 Apr 2009 13:22:46 +0000 (15:22 +0200)]
HAHA
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Thu, 23 Apr 2009 20:09:49 +0000 (22:09 +0200)]
p-u
check for the existance of Description: in a binary package.
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Sun, 19 Apr 2009 20:06:25 +0000 (22:06 +0200)]
Remove an unexisting package
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Sun, 19 Apr 2009 16:45:29 +0000 (18:45 +0200)]
which_apt_conf_file
do the same for the apt.conf files
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Sun, 19 Apr 2009 16:41:01 +0000 (18:41 +0200)]
which_conf_file
we should not return the path, but just set the Cnf values to whatever is
in that local config file, so that can point to the right dak.conf then.
and, as that will be different from default "/etc/dak/dak.conf" then let
the usual process read in our different config files
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Sun, 19 Apr 2009 16:37:46 +0000 (18:37 +0200)]
which_conf_file
if local config files are allowed, return the path for them, in case
one exists.
if not just behave as usual.
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Sun, 19 Apr 2009 16:15:04 +0000 (18:15 +0200)]
dak.conf-etc
add new file which really is etc/dak/dak.conf
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Sun, 19 Apr 2009 15:45:50 +0000 (17:45 +0200)]
eeks
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Sun, 19 Apr 2009 15:45:22 +0000 (17:45 +0200)]
also tell flotow, not only merkel
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Sat, 18 Apr 2009 21:59:41 +0000 (23:59 +0200)]
Merge branch 'merge'
* merge:
skip arches we should skip
i think contents are actually working on separate thread now?
possibly working multithread contents writer that has udebs disabled
change reject email template to solicit a reply if new files are uploaded
convert binaries.type to a enum
some tweaking to the contents generation queries
strip ./ from front of paths during contents.bootstrap
Joerg Jaspert [Sat, 18 Apr 2009 21:59:20 +0000 (23:59 +0200)]
Merge commit 'stew/content_generation' into merge
* commit 'stew/content_generation':
skip arches we should skip
i think contents are actually working on separate thread now?
possibly working multithread contents writer that has udebs disabled
change reject email template to solicit a reply if new files are uploaded
convert binaries.type to a enum
some tweaking to the contents generation queries
strip ./ from front of paths during contents.bootstrap
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Mark Hymers [Sat, 18 Apr 2009 16:28:16 +0000 (17:28 +0100)]
automatically clean transitions
Signed-off-by: Mark Hymers <mhy@debian.org>
Mark Hymers [Sat, 18 Apr 2009 16:23:51 +0000 (17:23 +0100)]
allow us to kill apt-ftparchive clean without rerunning from there
Signed-off-by: Mark Hymers <mhy@debian.org>
Mark Hymers [Sat, 18 Apr 2009 16:16:36 +0000 (17:16 +0100)]
:
Signed-off-by: Mark Hymers <mhy@debian.org>
Mark Hymers [Sat, 18 Apr 2009 16:13:28 +0000 (17:13 +0100)]
fix variable name and send email before removing transition
Signed-off-by: Mark Hymers <mhy@debian.org>
Mark Hymers [Sat, 18 Apr 2009 14:29:34 +0000 (15:29 +0100)]
merge transitions work
Signed-off-by: Mark Hymers <mhy@debian.org>
Mark Hymers [Sat, 18 Apr 2009 14:26:31 +0000 (15:26 +0100)]
optionally send emails on completed transitions
Signed-off-by: Mark Hymers <mhy@debian.org>
Mark Hymers [Sat, 18 Apr 2009 12:28:21 +0000 (13:28 +0100)]
add automatic mode for cleanups
Signed-off-by: Mark Hymers <mhy@debian.org>
Mark Hymers [Sat, 18 Apr 2009 12:28:08 +0000 (13:28 +0100)]
change print_info to get_info so we can use it for emailing later
Signed-off-by: Mark Hymers <mhy@debian.org>
Joerg Jaspert [Thu, 16 Apr 2009 21:05:30 +0000 (23:05 +0200)]
disallow deletion of notes for trainees
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Thu, 16 Apr 2009 20:58:24 +0000 (22:58 +0200)]
Also tell who locked something
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Thu, 16 Apr 2009 20:43:55 +0000 (22:43 +0200)]
Lets import the future
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Thu, 16 Apr 2009 20:31:56 +0000 (22:31 +0200)]
process-new
do a source-package based locking in NEW processing.
This means multiple people should be able to work in NEW at the same
time (SHOCKING NEWS).
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Wed, 15 Apr 2009 21:25:15 +0000 (23:25 +0200)]
And this should get rid of suddenly empty prod editors. I hope.
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Wed, 15 Apr 2009 21:20:46 +0000 (23:20 +0200)]
*sigh*
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Wed, 15 Apr 2009 21:20:00 +0000 (23:20 +0200)]
Why oh why doesnt this code just read my mind and adapt how i want it to be?
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Wed, 15 Apr 2009 21:17:03 +0000 (23:17 +0200)]
prod_maintainer wants to take arguments now
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Wed, 15 Apr 2009 21:14:58 +0000 (23:14 +0200)]
Oh my, all for well tested patches
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Wed, 15 Apr 2009 21:12:37 +0000 (23:12 +0200)]
not only does this want correct syntax, but also correct indexes for our tuples. bah.
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Wed, 15 Apr 2009 21:11:47 +0000 (23:11 +0200)]
Lets try it with correct syntax
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Wed, 15 Apr 2009 21:08:41 +0000 (23:08 +0200)]
process-new
notes now have a timestamp column.
do not allow to edit old notes, only add new ones
hand over existing notes (if any) to the reject message (unless we get called with -m)
hand over existing notes (if any) to the prod message
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Wed, 15 Apr 2009 17:48:58 +0000 (19:48 +0200)]
pn
do not try to close the log in trainee mode
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Wed, 15 Apr 2009 08:16:19 +0000 (10:16 +0200)]
p-n
yay, we want a string here
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Tue, 14 Apr 2009 23:32:05 +0000 (01:32 +0200)]
process-new
add a trainee mode which forbids certain actions to be taken.
This is turned on by the -t option or (more likely) by just being unable
to open the dak logfile.
In trainee mode one can edit and remove notes, "Check" a package and skip or quit
the tool, nothing else. Every other menu option just rewrites the menu.
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Tue, 14 Apr 2009 23:05:51 +0000 (01:05 +0200)]
pn
nicer display of existing notes
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Tue, 14 Apr 2009 22:59:40 +0000 (00:59 +0200)]
Spelling
yesyes python, you want them spelled correctly
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Tue, 14 Apr 2009 22:57:38 +0000 (00:57 +0200)]
Fixes
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Tue, 14 Apr 2009 22:56:32 +0000 (00:56 +0200)]
schema 11
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Tue, 14 Apr 2009 22:55:45 +0000 (00:55 +0200)]
NEW
stop storing NEW comments in .dak files and instead put them into the
database
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Tue, 14 Apr 2009 20:47:46 +0000 (22:47 +0200)]
process-unchecked
move files to NEW using 0640 (changes 0644) mode, not default 0660/0664
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Mike O'Connor [Tue, 14 Apr 2009 13:11:10 +0000 (09:11 -0400)]
skip arches we should skip
Mike O'Connor [Tue, 14 Apr 2009 12:23:51 +0000 (08:23 -0400)]
Merge branch 'master' into content_generation
Mike O'Connor [Tue, 14 Apr 2009 12:23:02 +0000 (08:23 -0400)]
i think contents are actually working on separate thread now?
Joerg Jaspert [Mon, 13 Apr 2009 20:20:10 +0000 (22:20 +0200)]
Merge commit 'djpig/djpig' into merge
* commit 'djpig/djpig':
cruft_report: Split NBS command by source package
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Frank Lichtenheld [Mon, 13 Apr 2009 19:44:13 +0000 (19:44 +0000)]
cruft_report: Split NBS command by source package
This makes the commands easier to check if there are a lot of
them or if some source packages lost a lot of binary packages-
Signed-off-by: Frank Lichtenheld <djpig@debian.org>
Mike O'Connor [Mon, 13 Apr 2009 20:09:23 +0000 (16:09 -0400)]
merge with master
Mike O'Connor [Mon, 13 Apr 2009 20:01:56 +0000 (16:01 -0400)]
possibly working multithread contents writer that has udebs disabled
Joerg Jaspert [Mon, 13 Apr 2009 15:48:40 +0000 (17:48 +0200)]
etch-m68k
disable uploads to m68k etch.
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Mon, 13 Apr 2009 14:08:59 +0000 (16:08 +0200)]
apt.conf
remove etch-m68k from apt.conf. No need to always regenerate its files.
There hasnt been an upload to it for months.
And starting today, etch-m68k is also untouchable. We should reject uploads for it
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Mark Hymers [Sat, 11 Apr 2009 22:53:08 +0000 (23:53 +0100)]
Merge commit 'mhy/master' into merge
Merge update10 changes
Signed-off-by: Mark Hymers <mhy@debian.org>
Mark Hymers [Sat, 11 Apr 2009 19:45:46 +0000 (20:45 +0100)]
src_uploaders deletions now cascade from source
Signed-off-by: Mark Hymers <mhy@debian.org>
Mark Hymers [Sat, 11 Apr 2009 19:38:04 +0000 (20:38 +0100)]
fix update script
Signed-off-by: Mark Hymers <mhy@debian.org>
Mark Hymers [Sat, 11 Apr 2009 18:36:03 +0000 (19:36 +0100)]
add update to fix up src_uploaders
Signed-off-by: Mark Hymers <mhy@debian.org>
Mark Hymers [Sat, 11 Apr 2009 18:35:06 +0000 (19:35 +0100)]
Merge commit 'mhyftpmaster/master'
Joerg Jaspert [Fri, 10 Apr 2009 14:40:48 +0000 (16:40 +0200)]
todo
fixed an item.
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Fri, 10 Apr 2009 14:39:21 +0000 (16:39 +0200)]
process-accepted
fix the mess in stable-install.
Its still not optimal, but it should no longer need manual changes just to do
an oldstable point release.
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Fri, 10 Apr 2009 13:31:16 +0000 (15:31 +0200)]
Various
Fix a number of epydoc errors we had.
Make dakdb/ have a much nicer looking api doc :)
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Thu, 9 Apr 2009 16:29:37 +0000 (18:29 +0200)]
Merge branch 'master' of ssh://ftp-master.debian.org/srv/ftp.debian.org/git/dak
* 'master' of ssh://ftp-master.debian.org/srv/ftp.debian.org/git/dak:
fix undefined cursor
Fix the is_dm check to deal with the fact we have typing in our DB layer
Joerg Jaspert [Thu, 9 Apr 2009 16:28:35 +0000 (18:28 +0200)]
TODO
add p-a oldstable mess
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Mark Hymers [Wed, 8 Apr 2009 22:19:52 +0000 (23:19 +0100)]
fix undefined cursor
Signed-off-by: Mark Hymers <mhy@debian.org>
Mark Hymers [Wed, 8 Apr 2009 20:53:08 +0000 (21:53 +0100)]
Revert "add suite wrapper class"
This reverts commit
a26b794c1190cacae572e4c416fd8b2450e4bd33 .
Conflicts:
daklib/dbconn.py
Signed-off-by: Mark Hymers <mhy@debian.org>
Mark Hymers [Wed, 8 Apr 2009 20:49:39 +0000 (21:49 +0100)]
Merge commit 'mhyftpmaster/merge'
Mark Hymers [Wed, 8 Apr 2009 20:48:29 +0000 (21:48 +0100)]
Merge commit 'ftpmaster/master'
Mark Hymers [Wed, 8 Apr 2009 20:45:46 +0000 (21:45 +0100)]
Fix the is_dm check to deal with the fact we have typing in our DB layer
This code is still fundamentally wrong (assuming unknown UIDs are DDs is
idiotic; of course we know that the keys have had to pass the sig check
against the keyring so it's not a security issue thankfully; it just
might give a one shot limited window for DMs to upload non-DM packages)
Signed-off-by: Mark Hymers <mhy@debian.org>
Mark Hymers [Wed, 8 Apr 2009 20:24:52 +0000 (21:24 +0100)]
Merge branch 'master' into merge
Mark Hymers [Wed, 8 Apr 2009 20:24:44 +0000 (21:24 +0100)]
Merge commit 'origin/master' into merge
Mark Hymers [Wed, 8 Apr 2009 20:24:13 +0000 (21:24 +0100)]
Merge commit 'origin/master'
Joerg Jaspert [Tue, 7 Apr 2009 22:56:31 +0000 (00:56 +0200)]
Merge commit 'djpig/djpig' into merge
* commit 'djpig/djpig':
cruft_report: Hack: skip debian-installer on kfreebsd
cruft_report: Fix syntax error
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Frank Lichtenheld [Tue, 7 Apr 2009 22:51:52 +0000 (22:51 +0000)]
cruft_report: Hack: skip debian-installer on kfreebsd
Doesn't exist yet.
Frank Lichtenheld [Tue, 7 Apr 2009 22:51:13 +0000 (22:51 +0000)]
cruft_report: Fix syntax error
Joerg Jaspert [Tue, 7 Apr 2009 22:47:46 +0000 (00:47 +0200)]
Merge branch 'master' of ssh://ftp-master.debian.org/srv/ftp.debian.org/git/dak
* 'master' of ssh://ftp-master.debian.org/srv/ftp.debian.org/git/dak:
set STAMP variable during dinstall (mhy)
Joerg Jaspert [Tue, 7 Apr 2009 22:47:33 +0000 (00:47 +0200)]
dinstall
also generate the arch-space/pkg-nums stats here
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Archive Administrator [Tue, 7 Apr 2009 11:54:40 +0000 (11:54 +0000)]
set STAMP variable during dinstall (mhy)
Raphael Hertzog [Tue, 7 Apr 2009 06:52:12 +0000 (08:52 +0200)]
Merge branch 'master' into bugfixes
Conflicts:
dak/init_db.py
Joerg Jaspert [Mon, 6 Apr 2009 20:47:45 +0000 (22:47 +0200)]
signed buildd incoming dir
moved all the code creating this into the common file, new function
cron.unchecked now just calls this function (and so sources the common file)
cron.dinstall also calls this function, immediately after process-accepted did run.
This should get the times where buildds get 403 on files down to at max. the runtime
of process-accepted.
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Sun, 5 Apr 2009 21:53:54 +0000 (23:53 +0200)]
dinstall
tell merkel that it should sync the dd accessible parts.
do that after the long run of apt-ftparchive clean, so we have a large
delay between the earlier "sync projectb" and this, so to not overload
merkel too much.
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Sun, 5 Apr 2009 21:26:32 +0000 (23:26 +0200)]
dinstall
name the pre dumps differently.
delete predumps after 2 days.
also compress the dumpall files
hardlink same files.
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Sun, 5 Apr 2009 21:25:51 +0000 (23:25 +0200)]
expire_dumps
really tell the full rules
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Mark Hymers [Sun, 5 Apr 2009 19:10:36 +0000 (20:10 +0100)]
Merge branch 'master' into merge
Joerg Jaspert [Sun, 5 Apr 2009 16:30:39 +0000 (18:30 +0200)]
dinstall
move pgdakdev way later where it doesnt matter that it takes long to run
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Sun, 5 Apr 2009 16:30:15 +0000 (18:30 +0200)]
dinstall
actually call the stage function...
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Sun, 5 Apr 2009 13:16:51 +0000 (15:16 +0200)]
kfreebsd
add kfreebsd-[i386|amd64]
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Sat, 4 Apr 2009 13:48:42 +0000 (15:48 +0200)]
examine_package
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Sat, 4 Apr 2009 13:46:56 +0000 (15:46 +0200)]
examine_package
docstring
"clearup" relations variable after output.
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Sat, 4 Apr 2009 13:42:09 +0000 (15:42 +0200)]
examine_package
docstring
"clearup" relations variable after output.
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Sat, 4 Apr 2009 13:35:02 +0000 (15:35 +0200)]
process-new
and also call output_package_relations from process-new
Signed-off-by: Joerg Jaspert <joerg@debian.org>
Joerg Jaspert [Sat, 4 Apr 2009 13:24:22 +0000 (15:24 +0200)]
examine_package
output a list of package relations at the very end of our run.
should make it easier to compare them for packages that split into MANY
Signed-off-by: Joerg Jaspert <joerg@debian.org>