Github user JamesRTaylor commented on a diff in the pull request: https://github.com/apache/phoenix/pull/156#discussion_r58483157 --- Diff: phoenix-core/src/main/java/org/apache/phoenix/execute/ScanPlan.java --- @@ -209,7 +209,11 @@ protected ResultIterator newIterator(ParallelScanGrouper scanGrouper) throws SQL if (isOrdered) { scanner = new MergeSortTopNResultIterator(iterators, limit, orderBy.getOrderByExpressions()); } else { - if ((isSalted || table.getIndexType() == IndexType.LOCAL) && ScanUtil.shouldRowsBeInRowKeyOrder(orderBy, context)) { + if ((isSalted || table.getIndexType() == IndexType.LOCAL) + && ScanUtil.shouldRowsBeInRowKeyOrder(orderBy, context) + || (table.getIndexType() == IndexType.LOCAL && (Bytes.compareTo( + scan.getStartRow(), this.context.getScan().getStartRow()) != 0 || Bytes + .compareTo(scan.getStopRow(), this.context.getScan().getStopRow()) != 0))) { --- End diff -- Why is a merge sort required for this new condition? Is this the case when the original start/stop row has changed? Is this purely for the chunked result iterator case? Let's document this and add a TODO to remove as the chunked result iterator is deprecated (and if you could consistently add this TODO for other parts of code only necessary due to this deprecated chunking mechanism, that'd be much appreciated).
--- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---