diff --git a/examples/spark-hbase-demo2.md b/examples/spark-hbase-demo2.md
new file mode 100644
index 0000000..2ddca6c
--- /dev/null
+++ b/examples/spark-hbase-demo2.md
@@ -0,0 +1,20 @@
+## create spark sql table map to existing hbase (only single column mapping to hbase rowkey is supported)
+(1) create table in hbase
+```
+create 'hbase10k', 'f'
+for i in '1'..'10000' do for j in '1'..'2' do put 'hbase10k', "row#{i}", "f:c#{j}", "#{i}#{j}" end end
+```   
+
+(2) relate 'test' from SparkOnHbase
+```
+CREATE TABLE spark10k(rowkey STRING, a INTEGER, b INTEGER, PRIMARY KEY (rowkey)) MAPPED BY (hbase10k, COLS=[a=f.c1, b=f.c2]);
+```
+
+(3) Query :
+```
+   // test count *
+   (1) select count(*) from spark10k
+
+   // test group by
+   (2) select avg(a), b from spark10k group by b
+```
\ No newline at end of file