Sean Haynes dc677f0f18 Use scheduleOrThrow in MergeTree data selector thread pool ... At the moment, the use of scheduleOrThrowOnError doesn't currently have a timeout. So if you reach a point of saturation and use all threads available in the global pool, threads block infinitely and lead to a deadlock. This changes that behaviour so that MergeTree data selector threads will have a timeout and return a "No threads available" exception to clients. Credit to Nikita Mikhaylov for the proposition here: https://github.com/ClickHouse/ClickHouse/pull/56431		2024-03-22 16:46:20 +00:00
..
configs
__init__.py
test.py