Overview

Sort rows by inserting them into Persistit B-Tree and then read out in order.

Behavior

The rows of the input stream are written into a B-Tree that orders rows according to the ordering specification. Once the input stream has been consumed, the B-Tree is traversed from beginning to end to provide rows of the output stream.

Performance

PersistitSorter generates IO dependent on the size of the input stream. This occurs mostly during the loading phase, (when the input stream is being read). There will be some IO when the loaded B-Tree is scanned, but this is expected to be more efficient, as each page will be read completely before moving on to the next one.

Memory Requirements

Memory requirements (and disk requirements) depend on the underlying configuration, primarily the buffer pool size, and concurrent load on the system.

                context, bindings
        );


        inputCursor.openTopLevel();
        try {
            PersistitSorter sorter = new PersistitSorter(context, bindings, inputCursor, rowType, ordering, API.SortOption.PRESERVE_DUPLICATES, tap);
            RowCursor sortedCursor = sorter.sort();
            sortedCursor.open();
            try {
                while(sortedCursor.next() != null) {
                    // None
                }

                               Cursor input,
                               RowType rowType,
                               API.Ordering ordering,
                               API.SortOption sortOption,
                               InOutTap loadTap) {
        return new PersistitSorter(context, bindings, input, rowType, ordering, sortOption, loadTap);
    }

Examples of com.foundationdb.qp.storeadapter.indexcursor.PersistitSorter

Overview

Behavior

Performance

Memory Requirements

Related Classes of com.foundationdb.qp.storeadapter.indexcursor.PersistitSorter