Depends on the number of users and requests I think. I run a small private instance with some old friends of mine. I have reduced the unicorn workers and the sidekiq concurrency. It works really well with a small amount of RAM.
Depends on the number of users and requests I think. I run a small private instance with some old friends of mine. I have reduced the unicorn workers and the sidekiq concurrency. It works really well with a small amount of RAM.