[
https://issues.apache.org/jira/browse/HBASE-25438?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17254252#comment-17254252
]
Michael Stack commented on HBASE-25438:
---------------------------------------
I tried adding:
{{diff --git a/dev-support/create-release/release-util.sh
b/dev-support/create-release/release-util.sh
index 5f7224fded..ba45a35ef3 100755
--- a/dev-support/create-release/release-util.sh
+++ b/dev-support/create-release/release-util.sh
@@ -369,6 +369,8 @@ function configure_maven {
MAVEN_OPTS="-Dorg.slf4j.simpleLogger.showDateTime=true
-Dorg.slf4j.simpleLogger.dateTimeFormat=HH:mm:ss ${MAVEN_OPTS}"
# Suppress gobs of "Download from central:" messages
MAVEN_OPTS="-Dorg.slf4j.simpleLogger.log.org.apache.maven.cli.transfer.Slf4jMavenTransferListener=warn
${MAVEN_OPTS}"
+ # Up retries and no pooling. We see connection resets that cause build fail
up on gcp at least.
+ MAVEN_OPTS=" -Dmaven.wagon.http.pool=false
-Dmaven.wagon.http.retryHandler.count=10 ${MAVEN_OPTS}"
MAVEN_LOCAL_REPO="${REPO:-$(pwd)/$(mktemp -d hbase-repo-XXXXX)}"
[[ -d "$MAVEN_LOCAL_REPO" ]] || mkdir -p "$MAVEN_LOCAL_REPO"
MAVEN_SETTINGS_FILE="${MAVEN_LOCAL_REPO}/tmp-settings.xml"}}
... to help improve the failure to download but doesn't seem to help. I am
seeing more of 'Connection reset'. See HBASE-25440
> Update create-release mvn in Dockerfile; its 3.6.0; make it 3.6.3
> -----------------------------------------------------------------
>
> Key: HBASE-25438
> URL: https://issues.apache.org/jira/browse/HBASE-25438
> Project: HBase
> Issue Type: Improvement
> Components: create-release
> Reporter: Michael Stack
> Priority: Major
> Fix For: 3.0.0-alpha-1
>
>
> Having trouble running RM builds from gcp. Downloads of artifacts are having
> connection reset. Apparently mvn 3.6.1 gets a better http wagon downloader.
> Doesn't seem to help in this particular case but let me get this in anyways.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
