#!/bin/bash
#
# Copyright 2009 Red Hat, Inc. and/or its affiliates.
# Released under the GPL
#
# Author:      Dan Kenigsberg <danken@redhat.com>
#
# ksmtuned - a simple script that controls whether (and with what vigor) ksm
# should search for duplicated pages.
#
# starts ksm when memory commited to qemu processes exceeds a threshold, and
# make ksm work harder and harder untill memory load falls below that
# threshold.
#
# send SIGUSR1 to this process right after a new qemu process is started, or
# following its death, to retune ksm accordingly
#
# needs testing and ironing. contact danken@redhat.com if something breaks.

if [ -f /etc/ksmtuned.conf ]; then
    . /etc/ksmtuned.conf
fi

debug() {
    if [ -n "$DEBUG" ]; then
        s="`/bin/date`: $*"
        [ -n "$LOGFILE" ] && echo "$s" >> "$LOGFILE" || echo "$s"
    fi
}


KSM_MONITOR_INTERVAL=${KSM_MONITOR_INTERVAL:-60}
KSM_NPAGES_BOOST=${KSM_NPAGES_BOOST:-300}
KSM_NPAGES_DECAY=${KSM_NPAGES_DECAY:--50}

KSM_NPAGES_MIN=${KSM_NPAGES_MIN:-64}
KSM_NPAGES_MAX=${KSM_NPAGES_MAX:-1250}
# millisecond sleep between ksm scans for 16Gb server. Smaller servers sleep
# more, bigger sleep less.
KSM_SLEEP_MSEC=${KSM_SLEEP_MSEC:-10}

KSM_THRES_COEF=${KSM_THRES_COEF:-20}
KSM_THRES_CONST=${KSM_THRES_CONST:-2048}

total=`awk '/^MemTotal:/ {print $2}' /proc/meminfo`
debug total $total

npages=0
sleep=$[KSM_SLEEP_MSEC * 16 * 1024 * 1024 / total]
[ $sleep -le 10 ] && sleep=10
debug sleep $sleep
thres=$[total * KSM_THRES_COEF / 100]
if [ $KSM_THRES_CONST -gt $thres ]; then
    thres=$KSM_THRES_CONST
fi
debug thres $thres

KSMCTL () {
    case x$1 in
        xstop)
            echo 0 > /sys/kernel/mm/ksm/run
            ;;
        xstart)
            echo $2 > /sys/kernel/mm/ksm/pages_to_scan
            echo $3 > /sys/kernel/mm/ksm/sleep_millisecs
            echo 1 > /sys/kernel/mm/ksm/run
            ;;
    esac
}

committed_memory () {
    # calculate how much memory is committed to running qemu processes
    local pidlist
    pidlist=$(pgrep -d ' ' -- '^qemu(-kvm|:.{1,11})$')
    if [ -n "$pidlist" ]; then
        ps -p "$pidlist" -o rsz=
    fi | awk '{ sum += $1 }; END { print sum }'
}

free_memory () {
    awk '/^(MemFree|Buffers|Cached):/ {free += $2}; END {print free}' \
                /proc/meminfo
}

increase_npages() {
    local delta
    delta=${1:-0}
    npages=$[npages + delta]
    if [ $npages -lt $KSM_NPAGES_MIN ]; then
        npages=$KSM_NPAGES_MIN
    elif [ $npages -gt $KSM_NPAGES_MAX ]; then
        npages=$KSM_NPAGES_MAX
    fi
    echo $npages
}


adjust () {
    local free committed
    free=`free_memory`
    committed=`committed_memory`
    debug committed $committed free $free
    if [ $[committed + thres] -lt $total -a $free -gt $thres ]; then
        KSMCTL stop
        debug "$[committed + thres] < $total and free > $thres, stop ksm"
        return 1
    fi
    debug "$[committed + thres] > $total, start ksm"
    if [ $free -lt $thres ]; then
        npages=`increase_npages $KSM_NPAGES_BOOST`
        debug "$free < $thres, boost"
    else
        npages=`increase_npages $KSM_NPAGES_DECAY`
        debug "$free > $thres, decay"
    fi
    KSMCTL start $npages $sleep
    debug "KSMCTL start $npages $sleep"
    return 0
}

function nothing () {
    :
}

loop () {
    trap nothing SIGUSR1
    while true
    do
        sleep $KSM_MONITOR_INTERVAL &
        wait $!
        adjust
    done
}

PIDFILE=${PIDFILE-/var/run/ksmtune.pid}
if touch "$PIDFILE"; then
  loop &
  echo $! > "$PIDFILE"
fi