Lost connection to MySQL server during about octopus HOT 5 CLOSED

This is not a elegant solution but I think it will work:

if defined?(PhusionPassenger)
  PhusionPassenger.on_event(:starting_worker_process) do |forked|
    if forked
      ActiveRecord::Base.connection_proxy.instance_variable_get(:@shards).each do |shard, connection_pool|
        connection_pool.disconnect!
      end
    end
  end
end

But I'm using puma these days which not require this kind of configuration.

Closing :)

from octopus.

This erro also happens on unicorn, and putting this line on the after_fork seems to fix it:

ActiveRecord::Base.connection_proxy.instance_variable_get(:@shards).each {|k,v| v.clear_reloadable_connections! }

Try putting this on the Passenger config. (I'm not sure how you should do this, but here is my guess):
if defined?(PhusionPassenger)
PhusionPassenger.on_event(:starting_worker_process) do |forked|
if forked
ActiveRecord::Base.connection_proxy.instance_variable_get(:@shards).each {|k,v| v.clear_reloadable_connections! }
end
end
end

from octopus.

FWIW, the fix that @tchandy mentioned above seems to be working for me. A similar fix is also in the StackOverflow post that the OP mentioned...look for the post from Jan 30 at the bottom.

If anyone else can confirm this, then this might be closed!

Edit: the fix that is working for me is the one for Passenger, I haven't tried Unicorn at all.

from octopus.

We also have been suffering from intermittent 'Lost connection' and 'ActiveModel::MissingAttributeError' errors.
After applying @sobrinho 's solution, those errors have gone completely.
Thanks so much!

from octopus.

😆

from octopus.