TL;DR

March Hare 2.0.0.rc3 is released to rubygems.org.

This release is a release candidate of 2.0 that introduces another usability improvement.

2.0 should be production ready from the stability perspective but has breaking API changes.

Changes Between 2.0.0-rc2 and 2.0.0-rc3

Consumer Work Pool Tuning Improvements

MarchHare 1.x used to maintain a separate executor (thread pool) per non-blocking consumer. This is not optimal and reimplements the wheel RabbitMQ Java client already has invented: it dispatches consumer methods in a thread pool maintained by every connection.

Instead of maintaining its own executor, MarchHare now relies on the Java client to do the job. The key difference is that 1.x versions used to maintain a thread pool per channel while 2.x has a thread pool per connection.

rc3 introduces an easier way to configure the executor: it is now sufficient to only provide pool size. For example:

1

MarchHare.connect(:thread_pool_size => 16)

It is still possible to override the executor when opening a connection by providing an executor factory (any Ruby callable):

1
2
3

MarchHare.connect(:executor_factory => Proc.new {
  MarchHare::ThreadPools.fixed_of_size(16)
})

Change Log

Full change log for 2.0 is available on GitHub.

Michael on behalf of the Ruby RabbitMQ Clients Team