NihalJain commented on code in PR #6435:
URL: https://github.com/apache/hbase/pull/6435#discussion_r1848688312
##########
hbase-mapreduce/src/test/java/org/apache/hadoop/hbase/mapreduce/TestRowCounter.java:
##########
@@ -524,6 +526,106 @@ public void testInvalidTable() throws Exception {
}
}
+ /**
+ * Step 1: Add 6 rows(row1, row2, row3, row4, row5 and row6) to a table.
Each row contains 1
+ * column family and 4 columns. Step 2: Delete a column for row1. Step 3:
Delete a column family
+ * for row2 and row4. Step 4: Delete all versions of a specific column for
row3, row5 and row6.
+ * <p>
+ * Case 1: Run row counter without countDeleteMarkers flag Step a: Validate
counter values.
+ * <p>
+ * Case 2: Run row counter with countDeleteMarkers flag Step a: Validate
counter values.
+ */
+ @Test
+ public void testRowCounterWithCountDeleteMarkersOption() throws Exception {
+ // Test Setup
+
+ final TableName tableName =
+ TableName.valueOf(TABLE_NAME + "_" + "withCountDeleteMarkersOption");
+ final byte[][] rowKeys = { Bytes.toBytes("row1"), Bytes.toBytes("row2"),
Bytes.toBytes("row3"),
+ Bytes.toBytes("row4"), Bytes.toBytes("row5"), Bytes.toBytes("row6") };
+ final byte[] columnFamily = Bytes.toBytes("cf");
+ final byte[][] columns =
+ { Bytes.toBytes("A"), Bytes.toBytes("B"), Bytes.toBytes("C"),
Bytes.toBytes("D") };
+ final byte[] value = Bytes.toBytes("a");
+
+ try (Table table = TEST_UTIL.createTable(tableName, columnFamily)) {
+ // Step 1: Insert rows with columns
+ for (byte[] rowKey : rowKeys) {
+ Put put = new Put(rowKey);
+ for (byte[] col : columns) {
+ put.addColumn(columnFamily, col, value);
+ }
+ table.put(put);
+ }
+ TEST_UTIL.getAdmin().flush(tableName);
+
+ // Steps 2, 3, and 4: Delete columns, families, and all versions of
columns
+ Delete deleteA = new Delete(rowKeys[0]).addColumn(columnFamily,
columns[0]);
+ Delete deleteB = new Delete(rowKeys[1]).addFamily(columnFamily);
+ Delete deleteC = new Delete(rowKeys[2]).addColumns(columnFamily,
columns[0]);
+ Delete deleteD = new Delete(rowKeys[3]).addFamily(columnFamily);
+ Delete deleteE = new Delete(rowKeys[4]).addColumns(columnFamily,
columns[0]);
+ Delete deleteF = new Delete(rowKeys[5]).addColumns(columnFamily,
columns[0]);
+
+ table.delete(deleteA);
+ table.delete(deleteB);
+ table.delete(deleteC);
+ table.delete(deleteD);
+ table.delete(deleteE);
+ table.delete(deleteF);
+ TEST_UTIL.getAdmin().flush(tableName);
+ }
+
+ RowCounter rowCounterWithoutCountDeleteMarkers = new RowCounter();
+ RowCounter rowCounterWithCountDeleteMarkers = new RowCounter();
+ rowCounterWithoutCountDeleteMarkers.setConf(TEST_UTIL.getConfiguration());
+ rowCounterWithCountDeleteMarkers.setConf(TEST_UTIL.getConfiguration());
+
+ // Invocation
+
+ rowCounterWithoutCountDeleteMarkers.run(new String[] {
tableName.getNameAsString() });
+ rowCounterWithCountDeleteMarkers
+ .run(new String[] { tableName.getNameAsString(), "--countDeleteMarkers"
});
+
+ // Validation
+
+ // Case 1:
+
validateCounterCounts(rowCounterWithoutCountDeleteMarkers.getMapReduceJob().getCounters(),
4, 0,
+ 0, 0, 0, 0, 0);
+
+ // Case 2:
+
validateCounterCounts(rowCounterWithCountDeleteMarkers.getMapReduceJob().getCounters(),
6, 30,
Review Comment:
> >We had just 4 rows right, now the job returns 6? it counts deleted rows
also now??
>
>
>
> All the rows will be returned when doing a raw scan.
Sure but is that how we want to define the tool to behave? Yes we are using
raw scan but thats an internal design decision, isnt it? we need to think from
the perspective of a user. I am good if we well define the tool behaviour in
release notes amd others are fine with this expectation.
> >Is not DELETE supposed to be 0, we did not delete any row right?
>
> DELETE_COLUMN should't be 4?
>
>
>
> Please refer to the [Delete
client](https://github.com/apache/hbase/blob/master/hbase-client/src/main/java/org/apache/hadoop/hbase/client/Delete.java).
It has the delete APIs and how they are matched to the delete marker types.
>
> addColumn -> Type.Delete
>
> addColumns -> Type.DeleteColumn
>
> addFamily -> Type.DeleteFamily
Ohk sure, I assume you have tested the scenarios well.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscr...@hbase.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
