[Piuparts-devel] Failed queue reruns

[Dave Steele]

On Sun, Nov 27, 2011 at 8:41 AM, Holger Levsen <holger at layer-acht.org> wrote:
> Hi Dave,
>
>... I've also manually deleted all successful
> logfiles starting with d, e and f today, b+c yesterday and a the day before.
> Also once more code is merged into "piatti", I will just delete all logs in
> sid and once thats retested, in wheezy.
>

Awesomeness!

>> If master could provide visibility into the size of his prioritized
>> work queues, ...
>
> my patch in branch
>  feature/master/state-command
> allows to access the package counts on master

Again, awesome, and a bit freaky. You're in my head.




As the dust settles on the re-run of the failed queue, please indulge
me once more:


$ ./piublocker
dependency failed -  1552
failed testing -  192

blocking free cum  rdeps package
 477      43  1509   497 sgml-data
 458      26  1058   477 docbook-xsl
 136     134   922   173 libreoffice-core
 121     121   801   165 iceweasel
  93      93   708    93 gnustep-base-common
  36      29   679    36 gforge-common
  35      34   645    35 gosa
  33      33   612    33 liquidsoap
  29      26   583   714 python-gi
  26      26   557    27 drupal6
  24      24   533    24 cyrus-common
 ...

d-f-t counts are down by half since October, and the percentage of
packages in s-t is close to an historic high.

But, the top 5 blockers on the list still represent half of the total
blocked. Three of those 5 are covered by a single bug/leftover
directory:

   http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=611418

The fruit is still low-hanging. I'll be working 611418 and python-gi today