README updated

Author: rgmacedo

README.md

@@ -259,11 +259,30 @@ PosixPassthrough Statistics
 
 ### Scalability test
 
+To test the data plane scalability, one can also resort to the `padll_scalability_bench` benchmark.
+The following configurations are only the changes with respect to the "Simple test" example.
+
+**PADLL configurations: options header**
+```yaml
+# include/padll/options/options.hpp
+- option_default_hsk_rules_file : "hsk-scale-test"
+```
+
+**PAIO configurations: options header**
+```yaml
+# include/paio/options/options.hpp
+- option_default_channel_differentiation_workflow : true
+- option_default_channel_differentiation_operation_type : false
+```
+
 ```shell
 $ cd /path/to/padll
-$ LD_PRELOAD=$PATH_PADLL/libpadll.so ./build/padll_scalability_bench
+$ vi benchmarking/bench.sh  # set the padll_path to the correct pathname
+# maximum number of threads is currently set to 128
+$ ./benchmarking/bench.sh Execute <number-of-stages> <number-of-threads/workflows>
 ```
 
+
 ### Connecting to the Cheferd control plane
 
 To execute with the [Cheferd](https://github.com/dsrhaslab/cheferd) control plane, one must set the following configurations at the options header file ([options.hpp](https://github.com/dsrhaslab/padll/blob/master/include/padll/options/options.hpp)). 

benchmarking/bench.sh

@@ -3,7 +3,7 @@
 # padll_path='/home1/07853/rgmacedo/padll/build'
 padll_path='path/to/padll/build'
 
-iops=100000000
+iops=10000000
 num_runs=1
 
 # $1 = number of stages
@@ -15,9 +15,9 @@ function Execute {
     export padll_workflows=$2
     for (( stage=1; stage<($1+1); stage++ )); do 
         LD_PRELOAD=$padll_path/libpadll.so $padll_path/padll_scalability_bench $num_runs $2 $iops $1  > /dev/null 2>&1 &
-    	echo "$?"
+    	# echo "$?"
     done
-    echo ""
+    echo ""; echo "Results are placed at /tmp/padll-scalability-results/."; echo ""; 
 }
 
 "$@"

files/hsk-simple-test

@@ -1,2 +1,2 @@
 1 create_channel 1000 posix 1000 open no_op
-2 create_object 1000 1 posix no_op no_op drl 10000 1000
+2 create_object 1000 1 posix no_op no_op drl 10000 20000