[ https://issues.apache.org/jira/browse/HIVE-22964?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17056824#comment-17056824 ]
Peter Vary commented on HIVE-22964: ----------------------------------- Hi Aditya Shah, * Yestus errors to fix: {code:java} ./ql/src/java/org/apache/hadoop/hive/ql/io/HiveInputFormat.java:490: , finalDirs = Collections.synchronizedList( new ArrayList<>());:12: warning: ',' is preceded with whitespace. ./ql/src/java/org/apache/hadoop/hive/ql/io/HiveInputFormat.java:490: , finalDirs = Collections.synchronizedList( new ArrayList<>());:56: warning: '(' is followed by whitespace. ./ql/src/java/org/apache/hadoop/hive/ql/io/HiveInputFormat.java:602: processForWriteIdsForMmRead(dir, conf, validWriteIdList, allowOriginals, finalPaths, pathsWithFileOriginals);: warning: Line is longer than 120 characters (found 121). {code} * The ASF related errors are not yours. * The way you handled the config deprecation seems ok to me. * Still would like to see more localized error handling: {code:java} try { [..] try { for (Future<Void> pathFuture : pathFutures) { pathFuture.get(); } } catch (InterruptedException | ExecutionException e) { for (Future<Void> future : pathFutures) { future.cancel(true); } throw new IOException(e); } } finally { [..] } {code} Do you strongly disagree, or just forgot? Keep submitting the fixed patch until we have a green run. Thanks, Peter > MM table split computation is very slow > --------------------------------------- > > Key: HIVE-22964 > URL: https://issues.apache.org/jira/browse/HIVE-22964 > Project: Hive > Issue Type: Improvement > Reporter: Aditya Shah > Assignee: Aditya Shah > Priority: Major > Attachments: HIVE-22964.1.patch, HIVE-22964.patch > > > Since for MM table we process the paths prior to inputFormat.getSplits() we > end up doing listing on the whole table at once. This could be optimized. -- This message was sent by Atlassian Jira (v8.3.4#803005)