[ https://issues.apache.org/jira/browse/BEAM-1190?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15766004#comment-15766004 ]
Eugene Kirpichov commented on BEAM-1190: ---------------------------------------- Dan - please note I'm suggesting not to ignore non-existent files entirely, but only to ignore *files that were yielded by glob match operation but reported as non-existent by per-file stat operation* - i.e. not include such files into bundles produced by splitIntoBundles; effectively this is just increasing precision of the glob matching. Can you elaborate on how this is dangerous? > FileBasedSource should ignore files that matched the glob but don't exist > ------------------------------------------------------------------------- > > Key: BEAM-1190 > URL: https://issues.apache.org/jira/browse/BEAM-1190 > Project: Beam > Issue Type: Bug > Components: sdk-java-core > Reporter: Eugene Kirpichov > Assignee: Eugene Kirpichov > > See user issue: > http://stackoverflow.com/questions/41251741/coping-with-eventual-consistency-of-gcs-bucket-listing > We should, after globbing the files in FileBasedSource, individually stat > every file and remove those that don't exist, to account for the possibility > that glob yielded non-existing files due to eventual consistency. -- This message was sent by Atlassian JIRA (v6.3.4#6332)