Github user JoshRosen commented on a diff in the pull request:
https://github.com/apache/spark/pull/2646#discussion_r18426749
--- Diff: python/pyspark/tests.py ---
@@ -152,7 +152,7 @@ def test_external_sort(self):
self.assertGreater(shuffle.DiskBytesSpilled, last)
def test_external_sort_in_rdd(self):
- conf = SparkConf().set("spark.python.worker.memory", "1m")
+ conf = SparkConf().set("spark.python.worker.memory", "10m")
--- End diff --
Why did this test originally change the worker memory? Is the goal here to
force spilling?
Maybe we could add an undocumented "always spill / always externalize"
configuration option to force spilling irrespective of memory limits in order
to test this code. I suppose that we still might want tests like this, though,
to check that the memory usage monitoring is working correctly, although I
suppose we could write a separate test that only tests the memory monitoring.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
