This is an automated email from the ASF dual-hosted git repository.
morningman pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/doris.git
The following commit(s) were added to refs/heads/master by this push:
new 7b417605e27 [fix](test) make hive compress split assertion BE-aware
(#60947)
7b417605e27 is described below
commit 7b417605e2769aded2b23082aff857280756ab79
Author: Socrates <[email protected]>
AuthorDate: Tue Mar 3 23:59:59 2026 +0800
[fix](test) make hive compress split assertion BE-aware (#60947)
### What problem does this PR solve?
Problem Summary:
`external_table_p0/hive/test_hive_compress_type.groovy` had a fixed
split assertion (`28`) for multi-BE, which is not stable across
environments.
The split behavior for COUNT pushdown depends on both `backendNum` and
`parallel_fragment_exec_instance_num`, so internal 3BE pipeline could
still produce `16` and fail.
This PR makes the assertion BE-aware and deterministic:
- read `backendNum` via `show backends`
- read `parallel_fragment_exec_instance_num` via `show variables`
- compute expected split count with the same condition as FE logic
- add comments to explain why this check differs between environments
---
.../hive/test_hive_compress_type.groovy | 33 +++++++++++++---------
1 file changed, 19 insertions(+), 14 deletions(-)
diff --git
a/regression-test/suites/external_table_p0/hive/test_hive_compress_type.groovy
b/regression-test/suites/external_table_p0/hive/test_hive_compress_type.groovy
index 3dcab8869f0..8668e148a63 100644
---
a/regression-test/suites/external_table_p0/hive/test_hive_compress_type.groovy
+++
b/regression-test/suites/external_table_p0/hive/test_hive_compress_type.groovy
@@ -24,8 +24,14 @@ suite("test_hive_compress_type",
"p0,external,hive,external_docker,external_dock
}
def backends = sql """show backends"""
- def backendNum = backends.size();
- logger.info("get backendNum:" + backendNum);
+ def backendNum = backends.size()
+ logger.info("get backendNum: ${backendNum}")
+ // `parallel_fragment_exec_instance_num` may be displayed as
+ // `deprecated_parallel_fragment_exec_instance_num` in newer branches.
+ def parallelExecInstanceRows = sql("show variables like
'%parallel_fragment_exec_instance_num%'")
+ assertTrue(parallelExecInstanceRows.size() > 0)
+ def parallelExecInstanceNum = (parallelExecInstanceRows[0][1] as
String).toInteger()
+ logger.info("get ${parallelExecInstanceRows[0][0]}:
${parallelExecInstanceNum}")
for (String hivePrefix : ["hive3"]) {
String hms_port = context.config.otherConfigs.get(hivePrefix +
"HmsPort")
@@ -41,18 +47,17 @@ suite("test_hive_compress_type",
"p0,external,hive,external_docker,external_dock
// table test_compress_partitioned has 6 partitions with different
compressed file: plain, gzip, bzip2, deflate
sql """set file_split_size=0"""
- if (backendNum == 1) {
- explain {
- sql("select count(*) from test_compress_partitioned")
- contains "inputSplitNum=16, totalFileSize=734675596,
scanRanges=16"
- contains "partition=8/8"
- }
- } else {
- explain {
- sql("select count(*) from test_compress_partitioned")
- contains "inputSplitNum=28, totalFileSize=734675596,
scanRanges=28"
- contains "partition=8/8"
- }
+ // COUNT pushdown split behavior depends on:
+ // totalFileNum < parallel_fragment_exec_instance_num * backendNum
+ // test_compress_partitioned currently has 16 files.
+ def expectedSplitNum = 16
+ if (backendNum > 1) {
+ expectedSplitNum = (16 < parallelExecInstanceNum * backendNum) ?
28 : 16
+ }
+ explain {
+ sql("select count(*) from test_compress_partitioned")
+ contains "inputSplitNum=${expectedSplitNum},
totalFileSize=734675596, scanRanges=${expectedSplitNum}"
+ contains "partition=8/8"
}
qt_q21 """select count(*) from test_compress_partitioned where
dt="gzip" or dt="mix""""
qt_q22 """select count(*) from test_compress_partitioned"""
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
