mirror of
https://github.com/honeymoose/OpenSearch.git
synced 2025-02-18 19:05:06 +00:00
By forking off the `SAME` pool tasks and executing them in random order, we are actually creating unrealisticc scenarios and missing the actual order of operations (whatever task that puts the task on the `SAME` queue will always run before the `SAME` queued task will be executed currently). Also, added caching for the executors. It doesn't matter much, but saves some objects and makes debugging a little easier because executor object ids make more sense.