zhztheplayer commented on code in PR #11373:
URL:
https://github.com/apache/incubator-gluten/pull/11373#discussion_r2694687337
##########
tools/gluten-it/common/src/main/scala/org/apache/gluten/integration/h/TpchSuite.scala:
##########
@@ -76,6 +76,9 @@ class TpchSuite(
override private[integration] def dataWritePath(): String = {
val featureFlags = dataGenFeatures.map(feature =>
s"-$feature").mkString("")
+ if (dataDir.startsWith("hdfs://")) {
+ return
s"$dataDir/$TPCH_WRITE_RELATIVE_PATH-$dataScale-$dataSource$featureFlags"
+ }
Review Comment:
ditto
##########
.github/workflows/velox_backend_x86.yml:
##########
@@ -181,6 +181,11 @@ jobs:
apt remove openjdk-11* -y
fi
ls -l
/root/.m2/repository/org/apache/arrow/arrow-dataset/15.0.0-gluten/
+ - name: Setup Apache Hadoop/HDFS
+ if: matrix.os == 'ubuntu:22.04' && matrix.spark == 'spark-3.5' &&
matrix.java == 'java-17'
+ uses: beyondstorage/setup-hdfs@master
Review Comment:
Let's avoid `@master`. Consider sepecifying a SHA instead.
##########
.github/workflows/velox_backend_x86.yml:
##########
@@ -93,7 +93,7 @@ jobs:
- uses: actions/upload-artifact@v4
with:
name: velox-native-lib-centos-7-${{github.sha}}
- path: ./cpp/build/releases/
+ path: ./cpp/build/
Review Comment:
Would you mind explaining this change a bit?
##########
tools/gluten-it/common/src/main/scala/org/apache/gluten/integration/QueryRunner.scala:
##########
@@ -63,6 +65,13 @@ class QueryRunner(val source: String, val dataPath: String) {
Failure(query.id, e)
}
}
+
+ private def fileExists(datapath: String): Boolean = {
+ if (datapath.startsWith("hdfs:")) {
+ val uri = URI.create(datapath)
+ FileSystem.get(uri, new Configuration()).exists(new Path(uri.getPath))
+ } else new File(datapath).exists()
Review Comment:
It's fine to keep 2 branches here, but I would suggest doing a refactor
toward a unified approach in the next PR. Let me know your thoughts on this
@infvg
##########
tools/gluten-it/common/src/main/scala/org/apache/gluten/integration/ds/TpcdsSuite.scala:
##########
@@ -81,6 +81,9 @@ class TpcdsSuite(
"non_partitioned"
}
val featureFlags = dataGenFeatures.map(feature =>
s"-$feature").mkString("")
+ if (dataDir.startsWith("hdfs://")) {
+ return
s"$dataDir/$TPCDS_WRITE_RELATIVE_PATH-$dataScale-$dataSource-$partitionedFlag$featureFlags"
+ }
Review Comment:
ditto
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
