diff --git a/tune-profiles/throughput-performance/ktune.sh b/tune-profiles/throughput-performance/ktune.sh
new file mode 100755
index 0000000..6d4fbec
--- /dev/null
+++ b/tune-profiles/throughput-performance/ktune.sh
@@ -0,0 +1,113 @@
+#!/bin/sh
+
+VAR_SUBSYS_KTUNE="/var/lock/subsys/ktune"
+
+CPUSPEED_SAVE_FILE="/var/run/ktune-cpuspeed.save"
+CPUSPEED_ORIG_GOV="var/run/ktune-cpuspeed-governor.save"
+CPUSPEED_CFG="/etc/sysconfig/cpuspeed"
+CPUSPEED_INIT="/etc/init.d/cpuspeed"
+CPUS="/sys/devices/system/cpu*/cpufreq"
+
+THP_ENABLE="/sys/kernel/mm/redhat_transparent_hugepage/enabled"
+THP_SAVE="/var/run/ktune-thp.save"
+
+start() {
+	# Save currently enabled governor and/or cpuspeed config
+	cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_governor > $CPUSPEED_ORIG_GOV
+	if [ ! -e $CPUSPEED_SAVE_FILE -a -e $CPUSPEED_CFG ]; then
+		cp -p $CPUSPEED_CFG $CPUSPEED_SAVE_FILE
+		sed -e 's/^GOVERNOR=.*/GOVERNOR=performance/g' $CPUSPEED_SAVE_FILE > $CPUSPEED_CFG
+	fi
+
+	# Enable performance governor (best for maximum i/o throughput in most cases)
+	if [ -e $CPUSPEED_INIT ]; then
+		/sbin/service cpuspeed restart > /dev/null 2>&1
+	else
+		for cpu in $CPUS; do
+			echo performance > $cpu/scaling_governor
+		done
+	fi
+
+	# Make sure transparent hugepages are enabled
+	cut -f2 -d'[' $THP_ENABLE  | cut -f1 -d']' > $THP_SAVE
+	(echo always > $THP_ENABLE) > /dev/null 2>&1
+
+	return 0
+}
+
+stop() {
+	# Restore previous cpuspeed config
+	if [ -e $CPUSPEED_SAVE_FILE ]; then
+		cp -fp $CPUSPEED_SAVE_FILE $CPUSPEED_CFG
+		rm -f $CPUSPEED_SAVE_FILE
+	fi
+
+	# Re-enable previous governor
+	if [ -e $CPUSPEED_INIT ]; then
+		/sbin/service cpuspeed restart > /dev/null 2>&1
+	elif [ -e $CPUSPEED_ORIG_GOV ]; then
+		for cpu in $CPUS; do
+			echo $(cat $CPUSPEED_ORIG_GOV) > $cpu/scaling_governor
+		done
+	else
+		for cpu in $CPUS; do
+			echo userspace > $cpu/scaling_governor
+			cat $cpu/cpuinfo_max_freq > $cpu/scaling_setspeed
+		done
+	fi
+	if [ -e $CPUSPEED_ORIG_GOV ]; then
+		rm -f $CPUSPEED_ORIG_GOV
+	fi
+
+	# Restore transparent hugepages setting
+	if [ -e $THP_SAVE ]; then
+		(echo $(cat $THP_SAVE) > $THP_ENABLE) > /dev/null 2>&1
+		rm -f $THP_SAVE
+	fi
+
+	return 0
+}
+
+reload() {
+	start
+}
+
+status() {
+	return 0
+}
+
+case "$1" in
+    start)
+        [ -f "$VAR_SUBSYS_KTUNE" ] && exit 0
+        start
+        RETVAL=$?
+        ;;
+    stop)
+        [ -f "$VAR_SUBSYS_KTUNE" ] || exit 0
+        stop
+        RETVAL=$?
+        ;;
+    reload)
+        [ -f "$VAR_SUBSYS_KTUNE" ] && reload
+        RETVAL=$?
+        ;;
+    restart|force-reload)
+        [ -f "$VAR_SUBSYS_KTUNE" ] && stop
+        start
+        RETVAL=$?
+        ;;
+    condrestart|try-restart)
+        [ -f "$VAR_SUBSYS_KTUNE" ] || exit 0
+        stop
+        start
+        RETVAL=$?
+        ;;
+    status)
+        status
+        RETVAL=$?
+        ;;
+    *)
+        echo $"Usage: $0 {start|stop|restart|condrestart|status}"
+        RETVAL=2
+        ;;
+esac
diff --git a/tune-profiles/throughput-performance/ktune.sysconfig b/tune-profiles/throughput-performance/ktune.sysconfig
index d2091b5..a49512b 100644
--- a/tune-profiles/throughput-performance/ktune.sysconfig
+++ b/tune-profiles/throughput-performance/ktune.sysconfig
@@ -2,7 +2,7 @@
 
 # This is the ktune sysctl file.  You can comment this out to prevent ktune
 # from applying its sysctl settings.
-#SYSCTL="/etc/sysctl.ktune"
+SYSCTL="/etc/sysctl.ktune"
 
 # Use *.conf files in the ktune configuration directory /etc/ktune.d.
 #   Value: yes|no,  default: yes
diff --git a/tune-profiles/throughput-performance/sysctl.ktune b/tune-profiles/throughput-performance/sysctl.ktune
index 080fd31..a19c621 100644
--- a/tune-profiles/throughput-performance/sysctl.ktune
+++ b/tune-profiles/throughput-performance/sysctl.ktune
@@ -1,51 +1,19 @@
-# ktune sysctl settings for EL 5 servers
+# ktune sysctl settings for rhel6 servers, maximizing i/o throughput
+#
+# Minimal preemption granularity for CPU-bound tasks:
+# (default: 1 msec#  (1 + ilog(ncpus)), units: nanoseconds)
+kernel.sched_min_granularity_ns = 10000000
 
-# 256 KB default performs well experimentally, and is often recommended by ISVs.
-net.core.rmem_default = 262144
-net.core.wmem_default = 262144
-
-# When opening a high-bandwidth connection while the receiving end is under
-# memory pressure, disk I/O may be necessary to free memory for the socket,
-# making disk latency the effective latency for the bandwidth-delay product
-# initially.  For 10 Gb ethernet and SCSI, the BDP is about 5 MB.  Allow 8 MB
-# to account for overhead, to ensure that new sockets can saturate the medium
-# quickly.
-net.core.rmem_max = 8388608
-net.core.wmem_max = 8388608
-
-# Allow a deep backlog for 10 Gb and bonded Gb ethernet connections
-net.core.netdev_max_backlog = 10000
-
-# Always have one page available, plus an extra for overhead, to ensure TCP NFS
-# pageout doesn't stall under memory pressure.  Default to max unscaled window,
-# plus overhead for rmem, since most LAN sockets won't need to scale.
-net.ipv4.tcp_rmem = 8192 87380 8388608
-net.ipv4.tcp_wmem = 8192 65536 8388608
-
-# Always have enough memory available on a UDP socket for an 8k NFS request,
-# plus overhead, to prevent NFS stalling under memory pressure.  16k is still
-# low enough that memory fragmentation is unlikely to cause problems.
-net.ipv4.udp_rmem_min = 16384
-net.ipv4.udp_wmem_min = 16384
-
-# Ensure there's enough memory to actually allocate those massive buffers to a
-# socket.
-net.ipv4.tcp_mem = 8388608 12582912 16777216
-net.ipv4.udp_mem = 8388608 12582912 16777216
-
-# Filesystem I/O is usually much more efficient than swapping, so try to keep
-# swapping low.  It's usually safe to go even lower than this on systems with
-# server-grade storage.
-vm.swappiness = 30
+# SCHED_OTHER wake-up granularity.
+# (default: 1 msec#  (1 + ilog(ncpus)), units: nanoseconds)
+#
+# This option delays the preemption effects of decoupled workloads
+# and reduces their over-scheduling. Synchronous workloads will still
+# have immediate wakeup/sleep latencies.
+kernel.sched_wakeup_granularity_ns = 15000000
 
 # If a workload mostly uses anonymous memory and it hits this limit, the entire
 # working set is buffered for I/O, and any more write buffering would require
 # swapping, so it's time to throttle writes until I/O can catch up.  Workloads
 # that mostly use file mappings may be able to use even higher values.
-vm.dirty_ratio = 50
-
-# Ensure there's always some easily-dropped pagecache if the system is under
-# memory pressure from cached files, since it's much faster to page back in than
-# swap.
-# Doesn't exist for 2.6.30++ anymore
-# vm.pagecache = 90
+vm.dirty_ratio = 40