Examples with TableSnapshotRegionSplit org.apache.hadoop.hbase.mapreduce.TableSnapshotInputFormat.TableSnapshotRegionSplit used on opensource projects

Example 1 with TableSnapshotRegionSplit

use of org.apache.hadoop.hbase.mapreduce.TableSnapshotInputFormat.TableSnapshotRegionSplit in project hbase by apache.

the class TestTableSnapshotInputFormat method verifyWithMockedMapReduce.

private void verifyWithMockedMapReduce(Job job, int numRegions, int expectedNumSplits, byte[] startRow, byte[] stopRow) throws IOException, InterruptedException {
    TableSnapshotInputFormat tsif = new TableSnapshotInputFormat();
    List<InputSplit> splits = tsif.getSplits(job);
    Assert.assertEquals(expectedNumSplits, splits.size());
    HBaseTestingUtility.SeenRowTracker rowTracker = new HBaseTestingUtility.SeenRowTracker(startRow, stopRow);
    for (int i = 0; i < splits.size(); i++) {
        // validate input split
        InputSplit split = splits.get(i);
        Assert.assertTrue(split instanceof TableSnapshotRegionSplit);
        // validate record reader
        TaskAttemptContext taskAttemptContext = mock(TaskAttemptContext.class);
        when(taskAttemptContext.getConfiguration()).thenReturn(job.getConfiguration());
        RecordReader<ImmutableBytesWritable, Result> rr = tsif.createRecordReader(split, taskAttemptContext);
        rr.initialize(split, taskAttemptContext);
        // validate we can read all the data back
        while (rr.nextKeyValue()) {
            byte[] row = rr.getCurrentKey().get();
            verifyRowFromMap(rr.getCurrentKey(), rr.getCurrentValue());
            rowTracker.addRow(row);
        }
        rr.close();
    }
    // validate all rows are seen
    rowTracker.validate();
}