You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@impala.apache.org by st...@apache.org on 2025/02/08 14:26:05 UTC
(impala) 02/03: IMPALA-13727: Fix wrong regex for hostname in test_multiple_blocks_mt_dop
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 4273bcb65dfbf09bcce475430d316334e7d894e1
Author: stiga-huang <hu...@gmail.com>
AuthorDate: Mon Feb 3 09:22:28 2025 +0800
IMPALA-13727: Fix wrong regex for hostname in test_multiple_blocks_mt_dop
The test uses r'host=(\S+:[0-9]*)' to match the hosts of all the
fragment instances in query profile, i.e.
host_list = re.findall(r'host=(\S+:[0-9]*)', result.runtime_profile)
This is able to find the instance name in lines like
Instance d843c27e276bfa7a:a86450f600000006 (host=name:27002):
and get "name:27002" from it.
However, the runtime profile we get right after fetching the query
result and before closing the query might have a section of total time
after the fragment instance host, e.g.
Instance d843c27e276bfa7a:a86450f600000006 (host=name:27002):(Total: 19.999ms, non-child: 0.000ns, % non-child: 0.00%)
The regex pattern will match "name:27002):Total:" in the string. If not
all the instances have this "Total" section, the instance names mismatch
and the test will fail. See the query profile attached in the JIRA as an
example. Not sure how this happens. Maybe it's not guaranteed that the
profile is ready immediately when the last result row is fetched. There
might be a delay for the coordinator to process the final update of the
profile from executors. I think it's ok to fix the test first.
This fixes the regex pattern to be r'\(host=([^:]+:[0-9]*)\)' so it can
exactly match the instance name.
Tests
- Verified the fix locally
Change-Id: If70daa344403f2ae8617bf86e7896c2bbfd9e736
Reviewed-on: http://gerrit.cloudera.org:8080/22457
Reviewed-by: Impala Public Jenkins <im...@cloudera.com>
Tested-by: Impala Public Jenkins <im...@cloudera.com>
---
tests/query_test/test_scanners.py | 14 ++++++++------
1 file changed, 8 insertions(+), 6 deletions(-)
diff --git a/tests/query_test/test_scanners.py b/tests/query_test/test_scanners.py
index c8b703381..85b53951e 100644
--- a/tests/query_test/test_scanners.py
+++ b/tests/query_test/test_scanners.py
@@ -877,19 +877,20 @@ class TestParquet(ImpalaTestSuite):
num_rows_read_list = re.findall(r'RowsRead: [0-9.K]* \(([0-9]*)\)',
result.runtime_profile)
# The extra fragment is the "Averaged Fragment"
- assert len(num_rows_read_list) == 7
- assert len(ranges_complete_list) == 7
+ assert len(num_rows_read_list) == 7, result.runtime_profile
+ assert len(ranges_complete_list) == 7, result.runtime_profile
# Extract the host for each fragment instance. The first is the coordinator
# fragment instance.
- host_list = re.findall(r'host=(\S+:[0-9]*)', result.runtime_profile)
- assert len(host_list) == 7
+ host_list = re.findall(r'\(host=([^:]+:[0-9]*)\)', result.runtime_profile)
+ assert len(host_list) == 7, result.runtime_profile
+ assert len(set(host_list)) == 3, result.runtime_profile
total_rows_read = 0
# Skip the Averaged Fragment; it comes first in the runtime profile.
for num_row_read in num_rows_read_list[1:]:
total_rows_read += int(num_row_read)
- assert total_rows_read == TOTAL_ROWS
+ assert total_rows_read == TOTAL_ROWS, result.runtime_profile
# Again skip the Averaged Fragment; it comes first in the runtime profile.
# With mt_dop 2, every backend will have 2 instances.
@@ -900,7 +901,8 @@ class TestParquet(ImpalaTestSuite):
ranges_per_host[host] = 0
ranges_per_host[host] += int(ranges_complete_list[i])
for host in ranges_per_host:
- assert ranges_per_host[host] == 2
+ assert ranges_per_host[host] == 2, ("ScanRangesComplete for " + host
+ + " should be 2 in profile:\n" + result.runtime_profile)
finally:
self.client.clear_configuration()