Added tar workload

Author: rohankadekodi

.gitignore

@@ -15,6 +15,7 @@ sqlite3-trace/.libs
 
 rsync/workload/
 rsync/rsync
+tar/workload/
 
 !ycsb/core
 !ycsb/tracerecorder

README.md

@@ -88,6 +88,7 @@ $ rm -rf /mnt/pmem_emul/*
 8. `sqlite3-trace/` contains SQLite3 source code
 9. `tpcc-sqlite/` contains TPCC source code
 10. `ycsb/` contains YCSB source code
+11. `tar/` contains tar source code
 
 The [Experiments
 page](https://github.com/utsaslab/SplitFS/blob/master/experiments.md)

experiments.md

@@ -44,6 +44,7 @@ We evaluate and benchmark on SplitFS using different application benchmarks like
 4. SQLite: `cd scripts/tpcc; ./compile_sqlite.sh; cd ../..` -- This will compile SQLite3
 5. TPCC: `cd scripts/tpcc; ./compile_tpcc.sh; cd ../..` -- This will compile TPCC workload
 6. rsync: `cd scripts/rsync; ./compile_rsync.sh; cd ../..` -- This will compile rsync
+7. tar: `cd scripts/tar; ./compile_tar.sh; cd ../..` -- This will compile tar
 
 Note: The <num_threads> argument in the compilation scripts performs the compilation with the number of threads given as input to the script, to improve the speed of compilation. 
 
@@ -54,6 +55,7 @@ Note: The <num_threads> argument in the compilation scripts performs the compila
 1. YCSB: `cd scripts/ycsb; ./gen_workloads.sh; cd ../..` -- This will generate the YCSB workload files to be run with LevelDB, because YCSB does not natively support LevelDB, and has been added to the benchmarks of LevelDB
 2. TPCC: `cd scripts/tpcc; ./gen_workload.sh; cd ../..` -- This will create an initial database on SQLite on which to run the TPCC workload
 3. rsync: `cd scripts/rsync/; sudo ./rsync_gen_workload.sh; cd ../..` -- This will create the rsync workload according to the backup data distribution as mentioned in the Paper
+4. tar: `cd scripts/tar/; sudo ./gen_workload.sh; cd ../..` -- This will create the tar workload as mentioned in the paper
 
 ---
 
@@ -62,6 +64,7 @@ Note: The <num_threads> argument in the compilation scripts performs the compila
 1. YCSB: `cd scripts/ycsb; ./run_ycsb.sh; cd ../..` -- This will run all the YCSB workloads on LevelDB (Load A, Run A-F, Load E, Run E) with `ext4-DAX, NOVA strict, NOVA Relaxed, PMFS, SplitFS-strict` 
 2. TPCC: `cd scripts/tpcc; ./run_tpcc.sh; cd ../..` -- This will run the TPCC workload on SQLite3 with `ext4-DAX, NOVA strict, NOVA Relaxed, PMFS, SplitFS-POSIX`
 3. rsync: `cd scripts/rsync; ./run_rsync.sh; cd ../..` -- This will run the rsync workload with `ext4-DAX, NOVA strict, NOVA Relaxed, PMFS, SplitFS-sync`
+4. tar: `cd scripts/tar; ./run_tar.sh; cd ../..` -- This will run the tar workload with `ext4-DAX, NOVA strict, NOVA Relaxed, PMFS, SplitFS-POSIX, SplitFS-sync, SplitFS-strict`
 
 ---
 

scripts/tar/Makefile

@@ -0,0 +1,40 @@
+# Run tests for splitfs
+
+# The current working directory
+CWD=$(shell pwd)
+# Root of the repo
+ROOT=$(CWD)/../..
+# SplitFS path
+SFS_PATH=/mnt/pmem_emul/
+
+all: splitfs.posix splitfs.sync splitfs.strict
+
+# Compile with posix specific flags
+splitfs.posix_compile:
+	export LEDGER_DATAJ=0 && \
+    export LEDGER_POSIX=1 && \
+	export LEDGER_TAR=1 && \
+    $(MAKE) -C ${ROOT}/splitfs clean && \
+    $(MAKE) -e -C ${ROOT}/splitfs
+
+# Compile with sync specific flags
+splitfs.sync_compile:
+	export LEDGER_DATAJ=0 && \
+    export LEDGER_POSIX=0 && \
+	export LEDGER_TAR=1 && \
+    $(MAKE) -C ${ROOT}/splitfs clean && \
+    $(MAKE) -e -C ${ROOT}/splitfs
+
+# Compile with strict specific flags
+splitfs.strict_compile:
+	export LEDGER_DATAJ=1 && \
+    export LEDGER_POSIX=0 && \
+	export LEDGER_TAR=1 && \
+    $(MAKE) -C ${ROOT}/splitfs clean && \
+    $(MAKE) -e -C ${ROOT}/splitfs
+
+splitfs.posix: splitfs.posix_compile
+
+splitfs.sync: splitfs.sync_compile
+
+splitfs.strict: splitfs.strict_compile

scripts/tar/compile_tar.sh

@@ -0,0 +1,14 @@
+#!/bin/bash
+
+set -x
+
+cur_dir=$(pwd)
+root_dir=`readlink -f ../..`
+tar_dir=$root_dir/tar
+
+cd $tar_dir
+make clean
+./configure
+make
+
+cd $cur_dir

scripts/tar/gen_workload.sh

@@ -0,0 +1,49 @@
+#!/bin/bash
+
+set -x
+
+cur_dir=$(pwd)
+root_dir=`readlink -f ../..`
+pmem_dir=/mnt/pmem_emul
+linux_tar=linux-4.18.10.tar.gz
+linux_dir=linux-4.18.10
+
+cd $root_dir
+rm $linux_tar
+wget https://cdn.kernel.org/pub/linux/kernel/v4.x/linux-4.18.10.tar.gz
+
+mkdir -p $pmem_dir/repo
+cp $root_dir/$linux_tar $pmem_dir/repo
+cd $pmem_dir/repo
+
+for i in {1..10}
+do
+    mkdir folder$i
+done
+
+tar -xf $linux_tar
+for i in {1..5}
+do
+    cp -r $linux_dir folder$i/
+done
+
+rm -rf $linux_dir
+
+for i in {6..10}
+do
+    cd folder$i
+    dd if=/dev/urandom of=temp1.txt bs=1M count=50
+    dd if=/dev/urandom of=temp2.txt bs=1M count=50
+    dd if=/dev/urandom of=temp3.txt bs=1M count=50
+    dd if=/dev/urandom of=temp4.txt bs=1M count=50
+    dd if=/dev/urandom of=temp5.txt bs=1M count=50
+    cd ..
+done
+
+cd $pmem_dir
+tar -zcf tar_workload.tar.gz repo
+rm -rf repo
+
+cd $cur_dir
+mkdir -p $root_dir/tar/workload
+cp -r $pmem_dir/tar_workload.tar.gz $root_dir/tar/workload/

scripts/tar/run_fs.sh

@@ -0,0 +1,71 @@
+#!/bin/bash
+
+if [ "$#" -ne 2 ]; then
+    echo "Usage: sudo ./run_fs.sh <fs> <run_id>"
+    exit 1
+fi
+
+set -x
+
+workload=tar
+fs=$1
+run_id=$2
+current_dir=$(pwd)
+tar_dir=`readlink -f ../../tar`
+workload_dir=$tar_dir/workload
+pmem_dir=/mnt/pmem_emul
+boost_dir=`readlink -f ../../splitfs`
+result_dir=`readlink -f ../../results`
+fs_results=$result_dir/$fs/$workload
+
+if [ "$fs" == "splitfs-posix" ]; then
+    run_boost=1
+elif [ "$fs" == "splitfs-sync" ]; then
+    run_boost=1
+elif [ "$fs" == "splitfs-strict" ]; then
+    run_boost=1
+else
+    run_boost=0
+fi
+
+ulimit -c unlimited
+
+echo Sleeping for 5 seconds ..
+sleep 5
+
+run_workload()
+{
+
+    echo ----------------------- TAR WORKLOAD ---------------------------
+
+    mkdir -p $fs_results
+    rm $fs_results/run$run_id
+
+    rm -rf $pmem_dir/*
+    cp $workload_dir/tar_workload.tar.gz $pmem_dir && sync
+
+    cd $pmem_dir
+
+    sleep 5
+
+    date
+
+    if [ $run_boost -eq 1 ]; then
+        time $boost_dir/run_boost.sh -p $boost_dir -t nvp_nvp.tree $tar_dir/src/tar -xf ./tar_workload.tar.gz 2>&1 | tee $fs_results/run$run_id
+    else
+        time $tar_dir/src/tar -xf $pmem_dir/tar_workload.tar.gz 2>&1 | tee $fs_results/run$run_id
+    fi
+
+    date
+
+    cd $current_dir
+
+    echo Sleeping for 5 seconds . .
+    sleep 5
+
+}
+
+
+run_workload
+
+cd $current_dir

scripts/tar/run_tar.sh

@@ -0,0 +1,56 @@
+#!/bin/bash
+
+current_dir=$(pwd)
+setup_dir=`readlink -f ../configs`
+pmem_dir=/mnt/pmem_emul
+
+run_tar()
+{
+    fs=$1
+    for run in 1 2 3
+    do
+        sudo rm -rf $pmem_dir/*
+        sudo taskset -c 0-7 ./run_fs.sh $fs $run
+        sleep 5
+    done
+}
+
+echo "-- ext4 DAX --"
+cd $setup_dir
+sudo ./nova_config.sh
+cd $current_dir
+sudo $setup_dir/dax_config.sh
+run_tar dax
+
+echo "-- NOVA Relaxed --"
+cd $setup_dir
+sudo ./nova_relaxed_config.sh
+cd $current_dir
+run_tar relaxed_nova
+
+echo "-- NOVA --"
+cd $setup_dir
+sudo ./nova_config.sh
+cd $current_dir
+run_tar nova
+
+echo "-- PMFS --"
+cd $setup_dir
+sudo $setup_dir/pmfs_config.sh
+cd $current_dir
+run_tar pmfs
+
+echo "-- SplitFS POSIX --"
+make splitfs.posix
+sudo $setup_dir/dax_config.sh
+run_tar splitfs-posix
+
+echo "-- SplitFS SYNC --"
+make splitfs.sync
+sudo $setup_dir/dax_config.sh
+run_tar splitfs-sync
+
+echo "-- SplitFS STRICT --"
+make splitfs.strict
+sudo $setup_dir/dax_config.sh
+run_tar splitfs-strict

splitfs/fileops_nvp.c

@@ -1643,7 +1643,7 @@ void nvp_free_dr_mmaps()
 	lfds711_queue_umm_cleanup( &qs, NULL );
 
 #if DATA_JOURNALING_ENABLED
-	
+
 	while( lfds711_queue_umm_dequeue(&qs_over, &qe_over) ) {
 		temp_free_pool_of_dr_mmaps = LFDS711_QUEUE_UMM_GET_VALUE_FROM_ELEMENT( *qe_over );
 		addr = temp_free_pool_of_dr_mmaps->start_addr;
@@ -1654,7 +1654,7 @@ void nvp_free_dr_mmaps()
 		char new_path[256];
 		sprintf(fd_str, "/proc/self/fd/%d", temp_free_pool_of_dr_mmaps->dr_fd);
 		file_name_size = readlink(fd_str, new_path, sizeof(new_path));
-		if (readlink(file_name_size == -1)
+		if (file_name_size == -1)
 			assert(0);
 		new_path[file_name_size] = '\0';
 
@@ -1673,7 +1673,7 @@ void nvp_free_dr_mmaps()
 	}
 
 #endif // DATA_JOURNALING_ENABLED
-	
+
 }
 
 void nvp_free_btree(unsigned long *root, unsigned long *merkle_root, unsigned long height, unsigned long *dirty_cache, int root_dirty_num, int total_dirty_mmaps)