Add benchmark_scripts directory

benchmark_scripts/tpch/copy_script.sh

@@ -0,0 +1,8 @@
+#!/bin/bash
+
+sudo cp tpch_duckdb.service /etc/systemd/system
+sudo cp tpch_duckdb_mem_limit.service /etc/systemd/system
+sudo cp tpch_hyper.service /etc/systemd/system
+sudo cp tpch_monetdb.service /etc/systemd/system
+sudo cp tpch_starrocks.service /etc/systemd/system
+sudo cp tpch_clickhouse.service /etc/systemd/system

benchmark_scripts/tpch/run_tpch.sh

@@ -0,0 +1,266 @@
+#!/bin/bash
+
+
+# Function to run dbgen script
+run_dbgen() {
+  scale=$1
+  destination=$2
+  run_dbgen_path=$3
+
+  scale_value = "$scale"
+
+  # Convert scale factor
+  #if [ "$scale" == "100mb" ]; then
+  #  echo "Creating data for SF $scale" 
+ #   scale_value=0.1
+  #elif [ "$scale" == "1gb" ]; then
+  #  echo "Creating data for SF $scale"
+  #  scale_value=1
+  #elif [ "$scale" == "10gb" ]; then
+  #  echo "Creating data for SF $scale"
+  #  scale_value=10
+  # elif [ "$scale" == "20gb" ]; then
+  #  echo "Creating data for SF $scale"
+  #  scale_value=20
+ # else
+  #  echo "Unsupported scale factor: $scale"
+  #  exit 1
+  #fi
+
+  #"$dbgen" $scale_value $destination
+  "${run_dbgen_path}/run_dbgen.sh" $scale_value $destination $run_dbgen_path
+}
+
+# Function to run python benchmark script
+run_benchmark() {
+  data_dir=$1
+  database_type=$2
+  db_path=$3 # Probably useless
+  python_script_path=$4
+  #Set of memory limits to try when the normal execution is killed due to OOM
+  memory_limits=(10)
+
+  # Save the current working directory
+  original_dir=$(pwd)
+
+  # Change the working directory to a new location
+  cd $python_script_path
+
+  # Perform operations in the new directory
+  echo "Now in: $(pwd)"
+  echo "$database_type"
+
+  # Determine the current service name
+  service_name="tpch.service"
+  memory_limit=$(systemctl show -p MemoryLimit --value $service_name)
+  # Check if the memory limit is set or is infinite
+  if [ "$memory_limit" == "infinity" ]; then
+  	echo "No memory limit is set for the service."
+  	memory_limit_bytes=0
+  else
+  	# Convert the memory limit from bytes to gigabytes (GB)
+  	memory_limit_gb=$(($memory_limit / (1024 * 1024 * 1024)))
+  fi
+  # Print the MemoryLimit
+  if [ "$memory_limit" == "infinity" ]; then
+  	echo "MemoryLimit for $service_name: unlimited"
+  else
+  	echo "MemoryLimit for $service_name: $memory_limit bytes ($memory_limit_gb GB)"
+  fi
+
+  # Empty the memory caches for cold start
+  sudo sh -c 'sync && echo 3 > /proc/sys/vm/drop_caches'
+
+  case $database_type in
+    "duckdb")
+      sudo sh -c 'sync && echo 3 > /proc/sys/vm/drop_caches'
+      echo "DuckDB is running"
+      python3 main.py --benchmark tpch --database duckdb --data $data_dir --db_path $data_dir
+      # Store the exit status of the python script in a variable
+      exit_status=$?
+      echo "DuckDB was terminated"
+      db_file="${data_dir}/tpch.db"
+      # Check if file exists
+      if [ -f "$db_file" ]; then
+      	# If file exists, remove it
+      	sudo rm "$db_file"
+    	echo "File $db_file has been removed."
+      fi
+      echo $exit_status
+      if [ "$exit_status" -ne 0 ]; then
+      	sudo sh -c 'sync && echo 3 > /proc/sys/vm/drop_caches'
+      	echo "Run DuckDB with memory limit through its own mechanism"
+      	python3 main.py --benchmark tpch --database duckdb --data $data_dir --db_path $data_dir --mem_limit $memory_limit_gb
+      	# Check if file exists
+      	if [ -f "$db_file" ]; then
+      	# If file exists, remove it
+      	sudo rm "$db_file"
+    	echo "File $db_file has been removed."
+      	fi
+      fi
+
+      #if [ "$?" -ne 0 ]; then
+      #  echo "DuckDB is OOM"
+      #  memory_info=$(free -m)
+        # Extract the available memory value (in megabytes)
+	#available_memory=$(echo "$memory_info" | awk 'NR==2{print $7}')
+	# Convert megabytes to gigabytes
+	#available_memory_gb=$(printf "%.0f" $(echo "scale=2; $available_memory / 1024" | bc))
+	#echo "Run DuckDB with all available memory $available_memory_gb"
+	#sudo sh -c 'sync && echo 3 > /proc/sys/vm/drop_caches'
+	#python3 main.py --benchmark tpch --database duckdb --data $data_dir --db_path $db_path
+	#python3 main.py --benchmark tpch --database duckdb --data $data_dir --db_path $db_path --mem_limit $available_memory_gb
+      #  for mem_limit in "${memory_limits[@]}"; do
+      #    echo "Run DuckDB with memory limit $mem_limit"
+      #    sudo sh -c 'sync && echo 3 > /proc/sys/vm/drop_caches'
+      #    python3 main.py --benchmark tpch --database duckdb --data $data_dir --db_path $db_path --mem_limit $mem_limit
+      #  done
+     # fi
+     echo "DuckDB just ended"
+      ;;
+    "hyper")
+      sudo sh -c 'sync && echo 3 > /proc/sys/vm/drop_caches'
+      echo "Hyper is running"
+      python3 main.py --benchmark tpch --database hyper --data $data_dir --db_path $data_dir
+      db_file="${data_dir}/tpch.hyper"
+      # Check if file exists
+      if [ -f "$db_file" ]; then
+      	# If file exists, remove it
+      	sudo rm "$db_file"
+    	echo "File $db_file has been removed."
+      fi
+      echo "Hyper just ended"
+      ;;
+    "monetdb")
+      sudo sh -c 'sync && echo 3 > /proc/sys/vm/drop_caches'
+      check_and_create_monetdb_files_dir "$db_path"
+      echo "MonetDB is running"
+      python3 main.py --benchmark tpch --database monetdb --data $data_dir --dbname tpch --dbfarm "${db_path}/monetdb_files"
+      echo "MonetDB just ended"
+      ;;
+    "starrocks")
+     #Define the path for Starrocks files !!!!!!!!!!!!!!
+      starrocks_path="/home/michalis/Documents/UofT/MSRG/Carbon/starrocks_files/StarRocks-3.1.4"
+      cd $starrocks_path
+       #Start Starrocks server
+      ./fe/bin/start_fe.sh --daemon
+      sleep 1
+      ./be/bin/start_be.sh --daemon
+      sleep 3
+      cd $python_script_path
+      sudo sh -c 'sync && echo 3 > /proc/sys/vm/drop_caches'
+      echo "Starrocks is running"
+      python3 main.py --benchmark tpch --database starrocks --ip 127.0.0.1 --data $data_dir --name tpch
+      cd $starrocks_path
+      #Stop Starrocks server
+      ./be/bin/stop_be.sh --daemon
+      ./fe/bin/stop_fe.sh --daemon
+      sleep 1
+      echo "Starrocks just ended"
+      # Delete data from database outside python script
+      sudo rm -rf "$starrocks_path/be/storage/data"/*
+      cd $python_script_path
+      ;;
+    *)
+      echo "Unsupported database type: $database_type"
+      exit 1
+      ;;
+  esac
+
+  # Change back to the original directory
+  cd "$original_dir"
+
+  # Perform operations in the original directory
+  echo "Back to: $(pwd)"
+
+}
+
+
+# Function to check and create MonetDB files directory
+check_and_create_monetdb_files_dir() {
+  directory_path=$1
+  monetdb_files_dir="${directory_path}/monetdb_files"
+
+  if [ ! -d "$monetdb_files_dir" ]; then
+    echo "Creating MonetDB files directory: $monetdb_files_dir"
+    mkdir "$monetdb_files_dir"
+  fi
+}
+
+
+# Function to check and create data directory
+check_and_create_data_dir() {
+  scale=$1
+  directory_path=$2
+  data_dir="${directory_path}/data_${scale}"
+
+  if [ ! -d "$data_dir" ]; then
+    echo "Creating data directory: $data_dir"
+    mkdir "$data_dir"
+  fi
+}
+
+
+# Main script
+# Define the path to the virtual environment !!!!!!!!!!!!!!!!!!!!!!
+venv_path="/home/michalis/Documents/UofT/MSRG/Carbon/codebase/olap_benchmark/venv/bin/activate"
+
+# Activate the virtual environment
+source "$venv_path"
+
+# Check if the correct number of arguments is provided
+if [ "$#" -ne 4 ]; then
+  echo "Usage: $0 <directory_path> <run_dbgen_script> <python_script> <data_generation_flag>" 
+  exit 1
+fi
+
+directory_path=$1 # Path where the data are located
+run_dbgen_path=$2 # Path where the dbgen is located
+python_script_path=$3 # Path where the python script is licated
+data_gen_flag=$4 # Flag to generate data
+
+# Define an array of scale factors
+#scale_factors=("100mb" "1gb" "10gb" "20gb")
+
+# Convert the environment variable string that contains the scale factors into an array
+IFS=',' read -r -a scale_factors <<< "$SCALE_FACTORS"
+
+# Define an array of database types
+#databases=("duckdb" "hyper" "monetdb" "starrocks")
+databases=("duckdb" "hyper" )
+
+#Enable sudo access for python script
+sudo chmod -R a+r /sys/class/powercap/intel-rapl
+
+#sudo echo -1000 > /proc/$$/oom_score_adj
+
+if [ "$4" == "true" ]; then
+  # Iterate over the scale factors
+  for scale_factor in "${scale_factors[@]}"; do
+    check_and_create_data_dir $scale_factor "$directory_path"
+    run_dbgen $scale_factor "${directory_path}/data_${scale}" $run_dbgen_path
+  done
+fi
+# Iterate over the databases
+for database_type in "${databases[@]}"; do
+  # Iterate over the scale factors
+  for scale_factor in "${scale_factors[@]}"; do
+    run_benchmark "${directory_path}/data_${scale_factor}gb" "$database_type" "$directory_path" $python_script_path
+    echo ""
+  done
+done
+
+# Save the current working directory
+original_dir=$(pwd)
+
+# Change the working directory to a new location
+cd $python_script_path
+cd "scripts"
+results_path="$python_script_path/core/benchmarks/tpch/results/csv/"
+#python3 plotResults.py --csv_path $results_path --benchmark tpch
+
+# Change back to the original directory
+cd "$original_dir"
+
+# Perform operations in the original directory
+echo "Back to: $(pwd)"