[ https://issues.apache.org/jira/browse/HADOOP-16921?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17090707#comment-17090707 ]
Steve Loughran commented on HADOOP-16921: ----------------------------------------- happy for you to take up. * anything related to mark/position in the AWS SDK is usually caused by upload failures and retries, usually very intermittent and so hard o track down. * I'd start with adding some verifyOpen() checks on this and related methods, and add a check for the argument being null. Maybe a null pointer came back from the bytebuffer factory in some failure mode > NPE in s3a byte buffer block upload > ----------------------------------- > > Key: HADOOP-16921 > URL: https://issues.apache.org/jira/browse/HADOOP-16921 > Project: Hadoop Common > Issue Type: Sub-task > Components: fs/s3 > Affects Versions: 3.3.0 > Reporter: Steve Loughran > Priority: Minor > > NPE in s3a upload when fs.s3a.fast.upload.buffer = bytebuffer -- This message was sent by Atlassian Jira (v8.3.4#803005) --------------------------------------------------------------------- To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: common-issues-h...@hadoop.apache.org