ubuntu-bengali-manual team mailing list archive

[Daniel Holbach <daniel.holbach@xxxxxxxxxx>]

Public bug reported:

The LoCo Directory gets more and more functions and more and more
external data find its way into the LD. We don't need to udpate all the
data in every cron run.

 - update-countries: once a day (even less would be fine)
 - update-languages: once a day (even less would be fine)
 - lpupdate:
   + get list of teams (every 10-15 minutes?)
   + get list of admins/owners (every 10-15 minutes?)
   + get statistics (bug 605651) (every 2 hours?)
   + get all team members (so we can appoint loco contacts) (every 30 minutes?)
 - etc. - there will be even more stuff in the future.

To make sure we don't have too many concurrent cronjob going on, I
suggest we kick off "./manage.py update" every 10 minutes. It itself
should keep track of which job needs to be run in which intervals and
let the jobs leave stamp files somewhere, so we can check when it was
last run.

** Affects: loco-directory
     Importance: Medium
         Status: Triaged


** Tags: data

** Changed in: loco-directory
   Importance: Undecided => Medium

** Changed in: loco-directory
       Status: New => Triaged

** Tags added: data

-- 
Cleverer 'update' mechanism
https://bugs.launchpad.net/bugs/620394
You received this bug notification because you are a member of Ubuntu
Bengali Manual, which is subscribed to loco-directory.

Status in LoCo Directory: Triaged

Bug description:
The LoCo Directory gets more and more functions and more and more external data find its way into the LD. We don't need to udpate all the data in every cron run.

 - update-countries: once a day (even less would be fine)
 - update-languages: once a day (even less would be fine)
 - lpupdate:
   + get list of teams (every 10-15 minutes?)
   + get list of admins/owners (every 10-15 minutes?)
   + get statistics (bug 605651) (every 2 hours?)
   + get all team members (so we can appoint loco contacts) (every 30 minutes?)
 - etc. - there will be even more stuff in the future.

To make sure we don't have too many concurrent cronjob going on, I suggest we kick off "./manage.py update" every 10 minutes. It itself should keep track of which job needs to be run in which intervals and let the jobs leave stamp files somewhere, so we can check when it was last run.