Donal McBreen

Yes I'd also expect to also see GETs reported. I think as well though that the SET counts are not accurate as they only get triggered by calls to `increment`...

If you do `mrsk app boot` or `mrsk app exec` I think you would want the most recently deployed version, not what you have locally. With https://github.com/mrsked/mrsk/pull/143/commits/1ed4a37da2b73125c37044d04e5a9a5ba0efd62a we pull the...

Superseded by https://github.com/basecamp/kamal/pull/828

I've not added a method on `ActiveRecord::QueryCache::ClassMethods` so you can do `Post.ignore_query_cache { ... }` but can do if we want that.

> > I've not added a method on ActiveRecord::QueryCache::ClassMethods so you can do Post.ignore_query_cache { ... } but can do if we want that. > > Wouldn't this be more...

@byroot - the problem is that any write clears the query cache on all pools. See [here](https://github.com/rails/rails/blob/eac48e95db9e30579d1c86308bdce7a5dd49c50e/activerecord/lib/active_record/connection_handling.rb#L244-L247).

Closing in favour of https://github.com/rails/rails/pull/50721

Thanks @jorgemanrubia! There isn't a Rails issue for this - one was raised on the SolidCache repo though (https://github.com/rails/solid_cache/issues/102)

@fractaledmind - we've been running ONCE Campfire (https://once.com/campfire) with this setting since it was released and it's worked well so, +1 for this. With deferred transactions, busy exceptions were a...

HI @mattbrictson, My reasoning for the change is that the current behaviour is not really fail fast - the parallel runner joins the threads in turn so if the second...