> Here is the problem: > - New agent fo_notify. Needs to be included in the Scheduler.conf > file. On an upgrade, most (if not all) of the time Scheduler.conf will > exist and so will not be updated. Without the update, when the > scheduler is restarted after the upgrade, it will fail. It can't find > fo_notify.
I think this is two problems. 1) The Scheduler.conf won't have the new agent in it. I think this is #343 http://bugs.linux-foundation.org/show_bug.cgi?id=343 right? 2) Are you saying the scheduler bombs out on restart because the self-test agent gives an error? That seems like a bug. I think either A) if an agent isn't listed in Scheduler.conf then we shouldn't test for it or B) even if we decide we do want to test all installed agents, they shouldn't fail just because they aren't listed in Scheduler.conf. nealk might have written it to complain if an agent wasn't listed, but I think we want to support the ability to not use agents, right? What do you think the behavior should be? There should probably be a separate bug in bugzilla for this. > Questions: [snip] See my mail to the above bug for possible answers to the rest of your questions. Scheduler.conf is a particularly tricky case. This is one of the reasons nealk and I hoped to move that info into the database and have the user make changes via the UI. It would still be good to do that, but I don't know who will do the work. -- Matt Taggart [email protected] _______________________________________________ fossology mailing list [email protected] http://fossology.org/mailman/listinfo/fossology
