This automatically collects a bunch of Ruby's GC-related metrics that
will come in handy while we tune the Unicorn workers. Some of theres
are:
* runtime.ruby.class_count
* runtime.ruby.gc.malloc_increase_bytes
* runtime.ruby.gc.total_allocated_objects
* runtime.ruby.gc.total_freed_objects
* runtime.ruby.gc.heap_marked_slots
* runtime.ruby.gc.heap_available_slots
* runtime.ruby.gc.heap_free_slots
* runtime.ruby.thread_count
Check https://docs.datadoghq.com/tracing/runtime_metrics/ruby/#data-collected for the complete list.
The cool thing is that
> Runtime metrics can be viewed in correlation with your Ruby services
This will enrich the quality of the reported traces for all web
frameworks in our stack. That means not only memcached as we did in
https://github.com/openfoodfoundation/openfoodnetwork/pull/4266 but
PostgreSQL and Rails too.
After enabling, the Trace Search & Analytics page populates which allows
us to search traces and add APM queries to dashboards.
