[impala] 02/03: IMPALA-11572: deflake test_mt_dop_skew_lpt part 3

[st...@apache.org]

This is an automated email from the ASF dual-hosted git repository.

stigahuang pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/impala.git

commit 5a433c1092cb641cfbdd1e152e7e0b7a1dc214f3
Author: Zoltan Borok-Nagy <bo...@cloudera.com>
AuthorDate: Wed Mar 1 15:26:54 2023 +0100

    IMPALA-11572: deflake test_mt_dop_skew_lpt part 3
    
    test_mt_dop_skew_lpt is still flaky. It fails most of the
    time on Ozone or when erasure coding is enabled. With this change
    we won't execute this test in such configurations.
    
    Also the skew threshold is decreased to 0.80.
    
    Change-Id: I73917ca2179199fd1cd7247000cd4b2f6d33a510
    Reviewed-on: http://gerrit.cloudera.org:8080/19564
    Reviewed-by: Impala Public Jenkins <im...@cloudera.com>
    Reviewed-by: Michael Smith <mi...@cloudera.com>
    Tested-by: Impala Public Jenkins <im...@cloudera.com>
---
 tests/query_test/test_scanners.py | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/tests/query_test/test_scanners.py b/tests/query_test/test_scanners.py
index be5a47578..566442542 100644
--- a/tests/query_test/test_scanners.py
+++ b/tests/query_test/test_scanners.py
@@ -383,6 +383,8 @@ class TestHdfsScannerSkew(ImpalaTestSuite):
 
   @SkipIfLocal.multiple_impalad
   @pytest.mark.execute_serially
+  @SkipIf.not_hdfs
+  @SkipIfNotHdfsMinicluster.tuned_for_minicluster
   def test_mt_dop_skew_lpt(self, vector, unique_database):
     """IMPALA-11539: Sanity check for MT scan nodes to make sure that the intra-node
        skew is mitigated. For intra-node scan range assignment we are using dynamic
@@ -390,7 +392,7 @@ class TestHdfsScannerSkew(ImpalaTestSuite):
        the items in the queue are ordered by scan sizes from largest to smallest, i.e.
        we are doing Longest-Processing Time (LPT) scheduling."""
     def count_intra_node_skew(profile):
-      SKEW_THRESHOLD = 0.85
+      SKEW_THRESHOLD = 0.80
       lines = [line.strip() for line in profile.splitlines() if "- BytesRead: " in line]
       assert len(lines) == 7  # Averaged fragment + 6 fragment
       bytes_read_array = []