Eileen

I've been thinking more about this and I don't think this is that helpful to add to the error. A statement invalid is rarely due to a bad connection setup,...

Ok so I took a look at this and it's is only reproducible with the combination of postgres, a shard name that's not `default`, and a migration that changes the...

I'm not sure there's an easy solution to this in Rails. The problem is the migration opens its own connection and transaction using role `:writing` and shard `:default`. But the...

I reopened this because it is still a bug, just not one with an easy fix. There's a feature I was planning to implement that I think will resolve this...

The entire point of building the 3-tier config was for apps to stop doing "#{Rails.env.to_sym}_name" when defining connections. 😕

Thank you!

@redox I just fixed a performance regression in sprockets-rails master dealing with the code mentioned in https://github.com/rails/sprockets-rails/issues/225#issuecomment-73813620 - can you see if that fixes what you're seeing?

Are you seeing the same in a vanilla application? I'm not seeing a slowdown between tag v2.1.4 and tag v2.2.3 (using https://github.com/eileencodes/perf_regression_sprockets) - I'm actually seeing a speed up between...

hey sorry I didn't respond earlier @redox I was at a conference. So from your app it looks like you're saying it happens when there are a ton of required...

@willkoehler can you check master and see if the issue still exists for you there? I fixed a perf issue in the area of code you mentioned in an earlier...