Merge lp:~stub/launchpad/cronscripts into lp:launchpad
| Status: | Merged | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Approved by: | Stuart Bishop on 2010-08-06 | ||||||||||||
| Approved revision: | no longer in the source branch. | ||||||||||||
| Merged at revision: | 11395 | ||||||||||||
| Proposed branch: | lp:~stub/launchpad/cronscripts | ||||||||||||
| Merge into: | lp:launchpad | ||||||||||||
| Diff against target: |
459 lines (+264/-14) 14 files modified
configs/testrunner/launchpad-lazr.conf (+1/-0) lib/canonical/config/schema-lazr.conf (+5/-1) lib/lp/bugs/doc/bugnotification-sending.txt (+1/-1) lib/lp/bugs/doc/checkwatches.txt (+1/-1) lib/lp/bugs/doc/sourceforge-remote-products.txt (+1/-1) lib/lp/registry/doc/teammembership.txt (+1/-0) lib/lp/services/scripts/base.py (+63/-3) lib/lp/services/scripts/tests/cronscripts.ini (+2/-0) lib/lp/services/scripts/tests/example-cronscript.py (+30/-0) lib/lp/services/scripts/tests/test_cronscript_enabled.py (+144/-0) lib/lp/soyuz/doc/buildd-slavescanner.txt (+2/-2) lib/lp/soyuz/scripts/tests/test_processupload.py (+6/-5) lib/lp/testing/tests/test_standard_test_template.py (+6/-0) lib/lp/translations/doc/poexport-request.txt (+1/-0) |
||||||||||||
| To merge this branch: | bzr merge lp:~stub/launchpad/cronscripts | ||||||||||||
| Related bugs: |
|
| Reviewer | Review Type | Date Requested | Status |
|---|---|---|---|
| Abel Deuring (community) | code | 2010-08-06 | Approve on 2010-08-06 |
|
Review via email:
|
|||
Commit Message
Enable and disable cronscripts using a configuration file.
Description of the Change
Steps towards Bug #607391.
Adds a config file controlling cronscripts. This config file allows them to be selectively or in bulk disabled from running.
| Robert Collins (lifeless) wrote : | # |
Thanks for working on this Stuart, its key to making rollouts safer
and more reliable.
I did have a small question though; our conf system is kindof like an
ini file already; how does using a regular ini file from within it
help?
-Rob
| Stuart Bishop (stub) wrote : | # |
On Fri, Aug 20, 2010 at 12:28 AM, Robert Collins
<email address hidden> wrote:
> Thanks for working on this Stuart, its key to making rollouts safer
> and more reliable.
>
> I did have a small question though; our conf system is kindof like an
> ini file already; how does using a regular ini file from within it
> help?
Currently, altering our existing config files is a heavyweight
operation. If we can fix this operational issue, we could collapse the
cron .ini file into the main config. I just took a lightweight
approach that we can expand on (scheduling shutdowns requested in the
bug for instance). I suspect we will stick with the .ini approach
until move to a zookeeper type system to control runtime
configuration.
The alternative I considered was using a separate PostgreSQL
configuration database, but after discussion with Gary we decided even
a separate PostgreSQL database doesn't provide the robustness we want
(and is probably NIH other zookeeper type systems).
--
Stuart Bishop <email address hidden>
http://
| Robert Collins (lifeless) wrote : | # |
I'm ok with an ini file; but would like to note that changing our
config is -only- heavyweight because we've chosen to have it so.
We can, in principle, have another config file referenced just like
you are here; I don't particularly care for that though.
I don't like the idea of another pg db either; zookeepr or something
may be fairly far away.
So I'm +0.5 or something on the ini; I think its a shame to have
*another* mechanism in play, but I'm keen on getting this work out
here; I think we should consider the presence of the two config
systems a techdebt bug and file it as such though.
-Rob

looks good