Conversation:
Notices
-
If you "kill -9" a process (or oom kills something whatever) is it expected that an application gracefully deal with that? The reason I ask is if you 'sudo -u postgres postmaster $STUFF' and then 'kill -9' the parent you get a couple of leftover processes. However, if you use systemctl, you don't get the leftover process. The problem lies when you use etcd (or consul or Zookee…
-
that's not what it means in this case. We are just using it for testing. There is no kill -15 in front of it. Leaving the orphaned processes really makes an HA solution impossible. I guess cleanly or graceful is really not what I am talking about. I just want if the parent goes down for the children to die too. Sure, there might be db corruption. That's not really the issue. T…
-
Thanks as always, though I'm not sure what that gets us. Indirectly, it made me realize we are making some assumptions about what oomkiller does, I think. I think half the team has touched this case, so I'm not 100% sure who knows what. Is the customer paranoid? I should probably not answer that. That said, ultimately this isn't *that* big of a deal. We aren't talking about da…
-
I wonder if turning off oomkiller is an option for them. Maybe considering the situation it is just best to let the server eat itself.
-
They are solving for a non-existent issue. The answer to the paranoia question is yes.
It seems though that the issue was introduced in between #patroni 1.6.0 and 2.0.1. Somehow everyone else on the team missed that email.
-
don't get me wrong, I've had the exact same thoughts, but I definitely think they are entering 100 guns for 5 people in a bunker territory.
Doug Whitfield Loadaverage
All Hacker Poesy content and data are available under the