BtbN
BtbN
Some things I've noticed: You rename the environment file from .env.production to stack.env, without also updating the documentation to reflect that new path. Is that intended? Why the change in...
Also, you are running two replicas of redis and sidekiq. Does that work just like that? I'd imagine that you at least need to somehow configure replication between the two...
So sidekiq might actually be fine, yeah. But everything I find online in regards to redis looks a lot more involved. Like, I keep finding setups similar to what's outlined...
Not at all. Neither of those two PRs fix the issue that /opt/mastodon/public/system does not exist in the container (since it's in .dockerignore and initially does not even exist in...
Rebasing this one on those others, or vice versa, should be fairly straight forward. There's almost no overlap despite them touching the same file.
The main potential future issue I see with this vs. the flags approach is that with cg_read_pid now just always returning zero-pids, it might cause a similar issue again in...
I'm not sure how WSL1 is involved in this, I haven't used that in ages. This is an issue on the latest WSL.
Just checked, latest WSL boots with cgroup_no_v1=all, so it's full cgroup v2. The layout in /sys/fs/cgroup also looks correct for that to me.