On Thu, 2013-05-16 at 18:56 +0200, Pierre-Yves Chibon wrote: > On Thu, 2013-05-16 at 10:48 -0600, Kevin Fenzi wrote: > > On Thu, 16 May 2013 09:04:54 +0200 > > Pierre-Yves Chibon <pingou@xxxxxxxxxxxx> wrote: > > > > > Hi all, > > > > > > Yesterday I announced an idea that I've had of using datagrepper to > > > announce package ownership changes [1]. > > > The feedback seem positive, so I change the script [2] to send the > > > email directly, so basically we can run it via a cron job easily. > > > > > > Do we want to have this script as a cron job in the infrastructure? > > > All it needs to run is argparse and requests. > > > The following question is how often should the cron run, this is > > > opened for discussion. > > > > We could. > > > > I floated the idea a while back of having some kind of 'oddjobs' type > > server to run things like this on, and/or some kind of app that lets > > users run a datagrepper query from cron and mail the results. > > > > For now, I guess we could just put this on bapp02? But that would take > > a freeze break. > > > > Or we could put it on datagrepper01? > > Either places are fine for me, whatever is the easiest to do and > maintain in the future. > We could also consider having it on datagrepper01 (which iirc is off > freeze) and move it to bapp02 after freeze, that would give us a little > more flexibility in case we face unexpected bugs in deploying the cron. Now that the freeze is over, should we proceed? The remaining questions were: where and how often? Suggestions? Pierre _______________________________________________ infrastructure mailing list infrastructure@xxxxxxxxxxxxxxxxxxxxxxx https://admin.fedoraproject.org/mailman/listinfo/infrastructure
