[ https://issues.apache.org/jira/browse/FLINK-16411?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17068986#comment-17068986 ]
Robert Metzger commented on FLINK-16411: ---------------------------------------- Yes, I was wondering the same. I think I'll reach out to the Google folks running the mirror. I have also experimented with using {{-Dhttp.keepAlive=false}} instead of {{-Dmaven.wagon.http.pool=false}} (as this was suggested in WAGON-486), but that made the situation even worse. My experiments with using caching for all tasks look quite promising. I'll run 5 builds over night to see what the failure rate is. > Maven central connection timeouts on Azure Pipelines > ---------------------------------------------------- > > Key: FLINK-16411 > URL: https://issues.apache.org/jira/browse/FLINK-16411 > Project: Flink > Issue Type: Bug > Components: Build System / Azure Pipelines > Reporter: Robert Metzger > Assignee: Robert Metzger > Priority: Critical > Labels: pull-request-available > Fix For: 1.11.0 > > Time Spent: 20m > Remaining Estimate: 0h > > Some test stages invoke maven again, where additional dependencies are > downloaded, sometimes failing the build. > This ticket is about using the Google mirror wherever possible. > Examples of failing tests: > - > https://dev.azure.com/rmetzger/Flink/_build/results?buildId=5882&view=logs&j=636f54dd-dda5-5b4b-f495-2d92ec493b6c&t=6c30efdf-a92a-5da3-9a6a-004c8552b2df > A failure looks like this: > {code} > [ERROR] Failed to execute goal on project flink-hadoop-fs: Could not resolve > dependencies for project org.apache.flink:flink-hadoop-fs:jar:1.11-SNAPSHOT: > Could not transfer artifact > org.apache.flink:flink-shaded-hadoop-2:jar:2.8.3-10.0 from/to central > (https://repo.maven.apache.org/maven2): GET request of: > org/apache/flink/flink-shaded-hadoop-2/2.8.3-10.0/flink-shaded-hadoop-2-2.8.3-10.0.jar > from central failed: Connection reset -> [Help 1] > {code} -- This message was sent by Atlassian Jira (v8.3.4#803005)