[17:32] <Necrosan> I have a binary (simple dumb worker) that I need to run X amount of times. Individual clients must restart if they die. Can upstart handle something like this?
[17:33] <Necrosan> If so, how should I be executing it? A for loop or something within my upstart script or do I need to make an actual intelligent script to do the monitoring of processes?
[17:33] <Necrosan> (Which I guess would negate the purpose of upstart entirely)
[18:39] <Necrosan> figured it out w/ the worker/workers model setup. pretty slick
[20:46] <AnrDaemon> Necrosan: You can write two jobs. One master (stater), one the actual worker that runs the binary.
[20:46] <Necrosan> yeah, that's what I ended up doing
[20:46] <Necrosan> worked a treat
[20:46] <AnrDaemon> It's actually in the cookbook, I just don't have it on hand right now.
[20:47] <Necrosan> I think I might have stumbled upon it after I had already did it.
[20:47] <Necrosan> Helps to scroll down all the way ;)
[20:47] <AnrDaemon> :D
[20:48] <AnrDaemon> After a number of situations like that, I got a habit of reading the manuals at least diagonally before using something new.
[20:49] <AnrDaemon> Helps get to know the available features. Even if you are not going to use them right now, it somehow gets stuck in the back of your head, and pop when you stubble upon new requirements.
[20:49] <Necrosan> ye
[20:49] <Necrosan> Writing C mostly, was a nice break 8-)
[20:50] <Necrosan> (same problems there re: manuals, but you usually don't notice them until you experience undefined behavior)
[20:51] <AnrDaemon> Undefined? More like unexpected.
[20:52] <Necrosan> Ya, prob the more appropriate word