Skip to content

Commit

Permalink
temp fix: disable destory_process_group (#5)
Browse files Browse the repository at this point in the history
When a broken world (process group) is detected, attempting to destroy
the process group causes the program to hang. We temporarily disable
this call to prevent this deadlock situation. We will revisit this later.
  • Loading branch information
myungjin authored May 24, 2024
1 parent b909bd6 commit d346215
Showing 1 changed file with 4 additions and 1 deletion.
5 changes: 4 additions & 1 deletion multiworld/world_manager.py
Original file line number Diff line number Diff line change
Expand Up @@ -173,7 +173,10 @@ def remove_world(self, world_name):
logger.debug(f"destory process group for {world_name}")
self.set_world(world_name)
del self._worlds[world_name]
dist.destroy_process_group()
# FIXME: calling destroy_process_group() here causes program hang.
# we need to find out a right timing/way to call this function.
# calling this function is temporarily disabled.
# dist.destroy_process_group()
logger.debug(f"done removing world {world_name}")

def set_world(self, world_name):
Expand Down

0 comments on commit d346215

Please sign in to comment.